An Ising machine (IM), as a type of analog computer tailored for tackling intractable combinatorial optimization problems, has attracted remarkable attention in recent years. In contrast to the blossoming field of bespoke IM hardware, developing metaheuristics from IMs remains largely uninvestigated. Here, we propose a physics-inspired evolutionary computation paradigm, termed the Ising optimization framework (IOF); it comprises a unique Ising algorithm and a hybrid annealing scheme, which together are well-suited for solving quadratic unconstrained binary optimization (QUBO) problems embedded in Ising system energy. The Ising algorithm leverages a set of iterated self-mapping functions to evolve an Ising-spin swarm, enabling efficient energy minimization in artificial Ising systems while mitigating detrimental chaos. Complementing the algorithm, a hybrid annealing scheme integrating singular value dropout, bifurcation control, and a nudging strategy, is devised to augment the overall optimization capacity. The effectiveness of the IOF is validated on various Ising and Max-cut problems with decision variables ranging from 625 to 5000 in number. In comparison to four major types of methods for solving QUBOs, including IM simulations, nature-inspired algorithms, a state-of-the-art heuristic, and the commercial solver Gurobi, the IOF consistently demonstrates notable optimization quality and computational efficiency. This paper provides a theoretical foundation and practical guidelines for bridging Ising-inspired approaches with evolutionary computation, offering an evolutionary perspective on Ising optimizations and suggesting a fertile avenue for future research and application.

Nature evaluates fitness through dynamic interactions between organisms and their environments, as exemplified by Darwinian evolution. Conversely, physical systems—such as those described by the Ising spin model—quantify a form of "fitness" via measurable macroscopic properties, notably energy. Drawing on this analogy and the principles of Evolutionary Computation (EC), we introduce EvoIsing, a physics-informed optimization method grounded in the Ising model’s energy minimization dynamics. By mapping combinatorial problem costs to binary variables and embedding them into the Ising system’s energy landscape via differential spin dynamics, EvoIsing exploits the natural tendency of analog spins to evolve toward low-energy states. Experiments across multiple binary optimization tasks demonstrate robust performance gains and reliability compared to traditional methods, underscoring the utility of energy-based dynamics for solving complex problems including Ising problem, maximum cut problem and traveling salesman problem. Notably, this implicit energy-based evaluation mechanism obviates explicit fitness and selection procedures typical in EC, thereby potentially mitigating issues like premature convergence and heavy computational overhead. Ultimately, this work advocates a broader synthesis of physics-informed models and evolutionary strategies, opening new avenues for efficient, nature-inspired optimization methods that transcend conventional fitness-driven paradigms.

Data centers account for significant global energy consumption and a carbon footprint. The recent increasing demand for edge computing and AI advancements drives the growth of data center storage capacity. Energy efficiency is a cost-effective way to combat climate change, cut energy costs, improve business competitiveness, and promote IT and environmental sustainability. Thus, optimizing data center energy management is the most important factor in the sustainability of the world. Power Usage Effectiveness (PUE) is used to represent the operational efficiency of the data center. Predicting PUE using Neural Networks provides an understanding of the effect of each feature on energy consumption, thus enabling targeted modifications of those key features to improve energy efficiency. In this paper, we have developed Bidirectional Gated Recurrent Unit (BiGRU) based PUE prediction model and compared the model performance with GRU. The data set comprises 52,560 samples with 117 features using EnergyPlus, simulating a DC in Singapore. Sets of the most relevant features are selected using the Recursive Feature Elimination with Cross-Validation (RFECV) algorithm for different parameter settings. These feature sets are used to find the optimal hyperparameter configuration and train the BiGRU model. The performance of the optimized BiGRU-based PUE prediction model is then compared with that of GRU using mean squared error (MSE), mean absolute error (MAE), and R-squared metrics.

Recently, domain adaptation has emerged as a powerful technique for on-site partial discharge (PD) condition assessment in gas-insulated switchgear (GIS). However, most existing methods face three major challenges: 1) relying on a single source domain for model development poses difficulties in effectively utilizing source domain samples with distribution differences; 2) limited condition assessment for unknown fault samples on-site, which faces distributional differences between multiple source domains; and 3) handling only a single task, which makes it challenging to generalize to multiple tasks simultaneously. To address these concerns, we propose a class alignment multisource domain adaptation network (CLMSDAN) for GIS PD condition assessment with unknown faults. First, a diversity feature extractor is developed to extract diverse features while addressing the negative transfer issue caused by knowledge differences by mining both interdomain and intradomain features, thus enabling the transfer of rich knowledge at multiple levels. Second, a novel multisource domain adaptation approach is employed from multiple perspectives to align distribution and distinguish between shared and unknown classes. Finally, a multiclassifier complementary strategy is introduced to recognize unknown faults, which automatically filters out source domain irrelevant class samples while distinguishing the contributions of different source domains to the target task. Experimental results show that CLMSDAN achieves 94.86% accuracy in diagnosis and 93.38% in severity assessment, outperforming baseline methods by over 10% in both tasks. This highlights its superior generalization and robustness across varying conditions and noise levels.

Condition assessment for gas-insulated switchgear (GIS), which are crucial component of power systems, involves three interrelated aspects, i.e., partial discharge (PD) diagnosis, localization, and severity assessment. However, existing methods for GIS PD condition assessment perform these aspects as separate tasks, ignoring the mutual influence among them and leading to inferior performance. To settle the abovementioned issue, we propose a multitask learning network (MTLN) for GIS PD condition assessment. First, a multitask network was developed, taking severity assessment as the main task and diagnosis and localization as parallel auxiliary tasks. This model not only facilitates the extraction of the coupling relationship between diagnosis and localization but also furnishes pertinent feature information for severity assessment. Second, to deploy the developed model to label-free scenarios on-site, a novel subdomain adaptation is established. The process of subdomain adaptation considers the alignment of both intraclass and interclass information, incorporating a secondary filtering mechanism to mitigate the issue of feature mismatch caused by incorrect pseudo labels. Experimental results show that the proposed MTLN not only offers diagnosis and location information for severity assessment but also facilitates the exploration of the coupling relationship between diagnosis and localization, thereby enhancing the performance of GIS PD condition assessment.

Fast charging technology is crucial for widespread electric vehicle (EV) adoption. To enhance efficiency and scalability, charging equipment manufacturers are shifting towards a modular architecture in fast charging stations (FCSs). This architecture features multiple converter modules and charging ports, allowing flexible power allocation through module-to-port assignment. However, it introduces challenges, particularly when ports operate with fewer modules, necessitating EV charging scheduling schemes to allocate limited FCS capacity while maintaining high quality-of-service (QoS). Traditional scheduling methods are ill-suited for modular FCS settings due to unique characteristics such as discrete module-to-port allocation, state-of-charge-dependent charge curves, and power ramp rate limits. This work proposes a social welfare-inspired EV scheduling framework for modular FCSs, using lexicographic optimization and receding horizon control. The framework includes a computationally efficient charge curve model based on sliding convex hulls and a mathematical model tailored for modular FCSs. The three-stage lexicographic model, derived from Rawlsian and Benthamite social welfare theories, accommodates customer preferences and EV characteristics for high QoS provision. A welfare score metric, adapted from social welfare theories, is also introduced for multi-faceted QoS assessment. Across ceteris paribus experiments, the proposed framework consistently outperforms three benchmark methods, with a margin of up to 34% in welfare scores over the second-best method. In a diverse set of randomized EV arrival scenarios, the framework enables a median welfare around 85%, outperforming the benchmarks by at least 7.8%, with statistical tests confirming its significance. Moreover, ramp rate violations are kept at a minimum, while the computational efficiency and scalability are verified.

Amidst the global shift towards the electrification of mass transportation, the effective long- and short-term planning of electric buses (e-buses) is gaining precedence due to challenges such as limited driving range, charging infrastructure requirements, charging costs, and battery lifetime. In this work, we propose a three-module modeling and optimization framework for strategic, tactical, and operational planning of multi-depot e-bus shuttle fleets by developing a series of mixed-integer programming models. The vehicle scheduling module determines e-bus deployment and trip assignments, ensuring feasible service while leveling energy consumption and mitigating battery degradation across the fleet. The charger deployment and charging planning module determines the number of chargers to deploy at depots and e-bus charging schedules to minimize life cycle costs. This module integrates an e-bus charging model that accounts for limited charger availability and practical considerations such as minimum charging duration, charger recovery period, and out-of-office hours, along with a neural network-based battery degradation model to minimize degradation costs and enable uniform battery aging. Finally, the online charging scheduling module updates the charging schedules to handle uncertainties in trip energy consumption. Case studies on a university campus shuttle e-bus network demonstrate a life cycle cost reduction of up to 38.2%, including savings on charger procurement, electricity, and battery degradation. Moreover, the proposed framework facilitates up to a 90.2% decrease in degradation costs and up to a 92.2% reduction in aging non-uniformity, maintaining cost optimality under uncertainties with a deviation of less than 1.5% compared to an oracle model in randomly generated scenarios.

In real-world multi-objective problems, each objective has a totally different scale. However, some frequently-used multi-objective evolutionary algorithms (MOEAs) have no objective space normalization mechanisms. The effect of objective space normalization on the performance of decomposition-based MOEAs (e.g., MOEA/D and NSGA-III) has already been examined for artificial test problems (e.g., DTLZ and WFG) in the literature. In this paper, we examine its practical usefulness for real-world multi-objective problems using various MOEAs. Our experimental results clearly show that objective space normalization is needed not only in decomposition-based MOEAs but also in hypervolume-based MOEAs. We also explain why objective space normalization is needed in these two types of MOEAs.

Real-world multiobjective optimization problems (MOPs) usually have conflicting and differently-scaled objectives. To deal with such problems, objective space normalization is widely used in multiobjective evolutionary algorithm (MOEA) design, especially, in the design of decomposition-based MOEAs. It has been demonstrated that uniformly-distributed solutions can be obtained for badly-scaled MOPs by decomposition-based MOEAs with objective space normalization. Recently, weight vector scaling has also been used for badly-scaled MOPs. In some studies, it was argued that weight vector scaling and objective space normalization are essentially the same when applied to decomposition-based MOEAs. In this paper, we theoretically and empirically show the relation between objective space normalization and weight vector scaling. Our results demonstrate that similarities and differences between the two methods depend on the choice of a scalarizing function. How the choice between normalization and weight vector scaling affects decomposition-based MOEAs with solution assignment mechanisms is also analyzed.

To efficiently implement the truck-drone collaborative logistics system, we introduce a multi-objective truck-drone collaborative routing problem with delivery and pick-up services (MCRP-DP). A truck collaborating with a fleet of drones serves three types of customers that require delivery, pick-up, and simultaneous delivery & pick-up services, respectively. Different from most of the existing studies where the drone visits only one customer in a flight, we allow the drone to serve another customer requiring pick-up service when it completes a delivery service. Meanwhile, we simultaneously optimize three objectives: transportation costs, waiting time of vehicles (i.e., truck and drone), and service reliability. To solve MCRP-DP, we propose an objective space decomposition-based multi-objective evolutionary algorithm with adaptive resource allocation (ODEA-ARA) In ODEA-ARA, an objective space decomposition strategy is used to maintain the diversity while an adaptive resource allocation strategy is designed to improve convergence. We design an ensemble of relative improvement and relative contribution to assist the resource allocation and a variable neighborhood Pareto local search integrating 7 problem-specific neighborhood structures to improve the solution. Extensive computational experiments are carried out to evaluate the performance of ODEA-ARA. The experimental results show that ODEA-ARA outperforms its competitors. Meanwhile, several useful managerial insights are presented.

Fast charging has the potential to address range anxiety and facilitate the adoption of electric vehicles (EVs). However, the distinct characteristics of public fast charging, such as short dwell times, diverse EV charge curves, and uncertain battery management systems (BMS), makes accurate coordination challenging while using traditional EV charging scheduling algorithms. This paper proposes a charge curve and BMS aware EV charging scheduling framework for capacity-constrained fast charging stations (FCSs) with a heterogeneous customer mix. The proposed framework involves a lexicographic mixed-integer optimization problem that is solved using a receding horizon scheme. The main contributions are: 1) an EV charge curve model that accurately reflects fast charging characteristics and can be efficiently incorporated into an optimization problem; 2) the introduction of two customer types to ensure high and fair quality of service (QoS) among customers with diverse preferences and EV specifications; 3) a four-criteria lexicographic optimization formulation that yields a pareto-optimal solution and provides additional benefits such as increased intermediate state-of-charge levels and reduced charge ramps; and 4) the adoption of a receding horizon scheme and a bounding heuristic that effectively addresses uncertainties associated with EV arrivals and BMS actions. Extensive case studies reveal that the proposed framework consistently outperforms the three benchmark methods by achieving a charge fairness improvement of up to 10.41% compared to the second best performer, while maintaining over 90% fairness score even under various BMS uncertainties and remaining computationally efficient for practical use in fast charging stations.

Normalization is an important algorithmic component for multiobjective evolutionary algorithms (MOEAs). Different normalization methods have been proposed in the literature. Recently, several studies have been conducted to examine the effects of normalization methods. However, the existing evaluation methods for investigating the effects of normalization are limited due to their drawbacks. In this paper, we discuss the limitations of the existing evaluation methods. A new metric has been proposed to facilitate the investigation of normalization methods. Our analysis clearly shows the superiority of the proposed metric over the existing methods. We also use the proposed metric to compare three popular normalization methods on problems with different Pareto front shapes.

A real-world multiobjective optimization problem (MOP) usually has differently scaled objectives. Objective space normalization has been widely used in multiobjective optimization evolutionary algorithms (MOEAs). Without objective space normalization, most of the MOEAs may fail to obtain uniformly distributed and well-converged solutions on MOPs with differently scaled objectives. Objective space normalization requires information on the Pareto front (PF) range, which can be acquired from the ideal and nadir points. Since the ideal and nadir points of a real-world MOP are usually not knowna priori, many recently proposed MOEAs tend to estimate and update the two points adaptively during the evolutionary process. Different methods to estimate ideal and nadir points have been proposed in the literature. Due to inaccurate estimation of the two points (i.e., inaccurate estimation of the PF range), objective space normalization may deteriorate the performance of an MOEA. Different methods have also been proposed to alleviate the negative effects of inaccurate estimation. This article presents a comprehensive survey of objective space normalization methods, including ideal point estimation methods, nadir point estimation methods, and different methods based on the utilization of the estimated PF range.

The main challenge in constrained multiobjective optimization problems (CMOPs) is to appropriately balance convergence, diversity and feasibility. Their imbalance can easily cause the failure of a constrained multiobjective evolutionary algorithm (CMOEA) in converging to the Pareto-optimal front with diverse feasible solutions. To address this challenge, we propose a dual-population-based evolutionary algorithm, named c-DPEA, for CMOPs. c-DPEA is a cooperative coevolutionary algorithm which maintains two collaborative and complementary populations, termedPopulation1andPopulation2. In c-DPEA, a novel self-adaptive penalty function, termedsaPF, is designed to preserve competitive infeasible solutions inPopulation1. On the other hand, infeasible solutions inPopulation2are handled using a feasibility-oriented approach. To maintain an appropriate balance between convergence and diversity in c-DPEA, a new adaptive fitness function, namedbCAD, is developed. Extensive experiments on three popular test suites comprehensively validate the design components of c-DPEA. Comparison against six state-of-the-art CMOEAs demonstrates that c-DPEA is significantly superior or comparable to the contender algorithms on most of the test problems.

The early and reliable detection of COVID-19 infected patients is essential to prevent and limit its outbreak. The PCR tests for COVID-19 detection are not available in many countries, and also, there are genuine concerns about their reliability and performance. Motivated by these shortcomings, this article proposes a deep uncertainty-aware transfer learning framework for COVID-19 detection using medical images. Four popular convolutional neural networks (CNNs), including VGG16, ResNet50, DenseNet121, and InceptionResNetV2, are first applied to extract deep features from chest X-ray and computed tomography (CT) images. Extracted features are then processed by different machine learning and statistical modeling techniques to identify COVID-19 cases. We also calculate and report the epistemic uncertainty of classification results to identify regions where the trained models are not confident about their decisions (out of distribution problem). Comprehensive simulation results for X-ray and CT image data sets indicate that linear support vector machine and neural network models achieve the best results as measured by accuracy, sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve (AUC). Also, it is found that predictive uncertainty estimates are much higher for CT images compared to X-ray images.

Objective space normalization is important since areal-world multiobjective problem usually has differently scaled objective functions. Recently, bad effects of the commonly used simple normalization method have been reported for the popular decomposition-based algorithm MOEA/D. However, the effects of recently proposed sophisticated normalization methods have not been investigated. In this paper, we examine the effectiveness of these normalization methods in MOEA/D. We find that these normalization methods can cause performance deterioration. We also find that the sophisticated normalization methods are not necessarily better than the simple one. Although the negative effects of inaccurate estimation of the nadir point are well recognized in the literature, no solution has been proposed. In order to address this issue, we propose two dynamic normalization strategies which dynamically adjust the extent of normalization during the evolutionary process. Experimental results clearly show the necessity of considering the extent of normalization.

To online estimate the power system sensitivity matrix considering mitigating data saturation, a series of multivariable least-squares (MLS) algorithms are proposed and compared, including the ordinary MLS (OMLS), the weighted MLS (WMLS), the memory-limited OMLS (ML-ORMLS), the memory-limited WRMLS (ML-WRMLS), and the memory-fading ML-WRMLS (MF-ML-WRMLS). Considering enhancing computational efficiency and accuracy by mitigating data saturation, the last three of them are specifically derived for sensitivity matrix online estimation using online-measured data. The effectiveness of the presented algorithms is verified and compared in the Nordic 32 system for voltage sensitivity matrix estimation. The results illustrate the prime algorithm in practice.

Prosumer Energy Management System (PEMS) is the system used for intelligent energy management for a house-hold with on-site photovoltaic (PV) system using the input data from various sources: smart meter, PV management system, a distribution transformer, dispatching unit of a home, real-time pricing from utility provider data center. An efficient PEMS should be able to provide demand response (DR) management in response to the prices, on-site energy generation, and electricity load to contribute to the improvement of overall electricity grid reliability and reduce the costs of a prosumer. This paper presents the Smart PEMS based on two components: seasonal decomposition long short-term memory (SD-LSTM) based forecasting system for predicting electricity prices and PV system energy generation combined with Q-Learning based home appliances scheduler.

The high penetration level of renewable energy is thought to be one of the basic characteristics of future smart grids. Wind power, as one of the most increasing renewable energy, has brought a large number of uncertainties into the power systems. These uncertainties would require system operators to change their traditional ways of decision-making. This article provides a comprehensive survey of computational intelligence techniques for wind power uncertainty quantification in smart grids. First, prediction intervals (PIs) are introduced as a means to quantify the uncertainties in wind power forecasts. Various PI evaluation indices, including the latest trends in comprehensive evaluation techniques, are compared. Furthermore, computational intelligence-based PI construction methods are summarized and classified into traditional methods (parametric) and direct PI construction methods (nonparametric). In the second part of this article, methods of incorporating wind power forecast uncertainties into power system decision-making processes are investigated. Three techniques, namely, stochastic models, fuzzy logic models, and robust optimization, and different power system applications using these techniques are reviewed. Finally, future research directions, such as spatiotemporal and hierarchical forecasting, deep learning-based methods, and integration of predictive uncertainty estimates into the decision-making process, are discussed. This survey can benefit the readers by providing a complete technical summary of wind power uncertainty quantification and decision-making in smart grids.

This paper proposes an alternate approach to extend the soft-switching range of full bridge converter without additional devices. The idea is to actively control the LV MOSFETs instead of using them just for synchronous rectification as implemented in conventional phase-shift known as, Phase Shift Full Bridge (PSFB). Dual Phase Shift Modulation (DPSM) utilizes an additional control variable to increase the circulating current in the converter which aids in ZVS turn on of voltage fed bridge devices. To improve the overall performance of the converter for entire operating range, a hybrid modulation is proposed where proposed DPSM is implemented at lighter loads to extend the soft-switching range and conventional PSFB is implemented at higher loads. With the proposed hybrid modulation, a smaller series inductance is enough for ZVS over the selected load range and hence, this reduces the duty cycle loss and the circulating current at higher loads. As the proposed modulation strategy is a software-only solution, there is no additional expense for active or passive components. Analysis, design and implementation of the proposed hybrid modulation has been discussed in the paper. An experimental prototype of 1 kW is developed and ZVS turn-on of all HV devices even at 6 % load with proposed modulation has been demonstrated. An improved efficiency of more than 92 % is demonstrated till 15 % load with proposed hybrid modulation.

A composite differential evolution, named C2oDE, has been recently proposed in the literature for solving constrained optimization problems. C2oDE is an extension of composite differential evolution, named CoDE, which is a state-of-the-art algorithm for bound constrained optimization problems. The three main features of C2oDE are use of - 1) a strategy pool which balances the trade-off between convergence and diversity, 2) a constraint handling strategy which combines epsilon-constraint handling technique and superiority of feasible solutions method, and 3) a re-start mechanism for improving performance on complex constrained optimization problems. In the original study, C2oDE was found to perform remarkably on problems of IEEE CEC 2006 and IEEE CEC 2010 test suite. In this paper, we investigate the performance of C2oDE on IEEE CEC 2017 constrained optimization problems. We observe that the performance of C2oDE deteriorates significantly on large number of 50 D and 100 D problems. We also investigate the efficacy of the re-start mechanism of C2oDE and observe that it fails to improve the performance of C2oDE on IEEE CEC 2017 benchmark problems.

This paper is a review of uncertainty modeling techniques used in smart grid studies. The literature dealing with uncertainty from various sources in smart grid is analyzed and presented. In a modern power grid, the risk may arise due to different reasons; in-termittent renewable energy sources, uncertain consumer reactions on demand response, driving patterns of electric vehicles, etc. The paper has two objectives. First is to bring out the trends in uncertainty handling techniques used in electrical power system problems, and second to introduce the scope of new risk processing techniques with the perspective of recent smart grid issues.

The distributed filtering problem is addressed in this paper for the discrete-time Takagi–Sugeno (T–S) fuzzy systems with incomplete measurements. The system under consideration includes various network-induced uncertainties, e.g., sensor saturation, quantization error, communication delay, and packet dropouts. Specifically, all these uncertainties are assumed to occur in a stochastic way. In addition, the measurement scheduling issue is also addressed such that only a portion of measurements are broadcasting due to the communication constraints. The main focus is on the design of distributed filters based on the information received locally and from the neighborhood such that the desired estimation performance is guaranteed in terms of the decay rate and disturbance attenuation level. Based on the Lyapunov stability theory, the existence condition for such filters is first proposed and the filter gain parameters are then determined by solving an optimization problem. A simulation example is finally presented to illustrate the effectiveness of the new filtering techniques.

Arterial traffic is the artery of urban transport and loads huge traffic pressure. In order to alleviate its traffic pressure effectively, a coordinated arterial traffic type-2 fuzzy logic control (FLC) method is proposed. First, arterial traffic flow model and evaluation index model are set up, in which the turning vehicles and lane length are given full consideration. The traditional queue spillover phenomenon in the traffic models can be prevented here. Second, aiming at the coordination and dynamic uncertainty problem in arterial traffic, a coordinated arterial traffic type-2 fuzzy coordination control method is put forward. It consists of two-layer type-2 fuzzy controller, the basic control layer and the coordination layer. The former allocates green time according to the traffic situation of each intersection, while the latter adjusts each intersection's green time on basis of the vehicles between the intersection and the downstream intersections for the purpose of enlarging green wave band. Finally, in order to configure the high-dimensional complex parameters of the coordinated two-layer type-2 FLC effectively, the parameters of membership function and the rules of the two controllers are optimized alternately by gravitational search algorithm. The simulation results verify the effectiveness of the proposed method from several aspects.

This paper is concerned with the asynchronous state estimation for a class of discrete-time switched complex networks with communication constraints. An asynchronous estimator is designed to overcome the difficulty that each node cannot access to the topology/coupling information. Also, the event-based communication, signal quantization, and the random packet dropout problems are studied due to the limited communication resource. With the help of switched system theory and by resorting to some stochastic system analysis method, a sufficient condition is proposed to guarantee the exponential stability of estimation error system in the mean-square sense and a prescribed performance level is also ensured. The characterization of the desired estimator gains is derived in terms of the solution to a convex optimization problem. Finally, the effectiveness of the proposed design approach is demonstrated by a simulation example.

The large-scale integration of renewable energy sources (RESs) at the distribution network level provides several benefits and opportunities for the distribution network as well as the transmission network, despite having some technical challenges such as intermittency handling, network voltage control, etc. Such RESs are typically accompanied with battery storage systems (BSSs) to mitigate effects of their intermittency. The RES-BSS systems can therefore be used to provide ancillary services to the transmission network while providing technical benefits to the distribution network operator (DNO) and economic benefits to the RES owners. Further, computational intelligence based techniques can provide a near-optimal solution within fewer iterations of the algorithm, which is a critical requirement in fulfilling ancillary service requests. Therefore, this paper presents a fully distributed computational intelligence based technique so that the distribution network could handle ancillary service requests by the transmission system operator (TSO) in real-time. Specifically, an agent is associated with each node of the distribution network and it utilizes computational intelligence and communication with nearby agents to achieve the solution. Simulation studies on a modified IEEE 30-node test system are shown to validate our aforementioned thesis.

Advanced metering infrastructure which is an integral component of smart homes has aided in tapping the potential of the residential sector for demand side management (DSM). DSM in smart homes focus mainly on some power-intensive appliances which affect the household load profile significantly. This paper proposes an intelligent appliance control (IAC) algorithm to monitor and control the daily operation of these power-intensive appliances using their simulated load models. The proposed algorithm employs differential evolution (DE) algorithm along with a DSM strategy to limit the smart household power consumption at every half an hour to an optimum limit. The paper demonstrates the ability of the proposed algorithm in minimizing the households' monthly electricity bill, maximizing the peak load reduction and minimizing the problem of distribution transformer overloading. The paper also focuses on studying the impacts of time of use (TOU) electricity pricing on residential customers' behavior. The simulation results indicate that TOU pricing augments the benefits of the proposed algorithm both at the residential level and the distribution transformer level.

In this paper, a unified differential evolution algorithm, named UDE, is presented for real parameter constrained optimization problems. The proposed UDE algorithm is inspired from some popular DE variants existing in the literature such as CoDE, JADE, SaDE, and ranking-based mutation operator. The primary feature of UDE lies in unifying the main idea of CoDE, JADE, SaDE, and ranking-based mutation. UDE uses three trial vector generation strategies and two parameter settings. At each generation, UDE divides the current population into two sub-populations. In the top sub-population, UDE employs all the three trial vector generation strategies on each target vector, just like in CoDE. For the bottom sub-population, UDE employs strategy adaptation, in which the trial vector generation strategies are periodically self-adapted by learning from their experiences in generating promising solutions in the top sub-population. Further, UDE utilizes a DE mutation strategy based local search operation. The constraints are handled in UDE using static penalty method. In contrast to most of the DE variants presented in the literature, UDE employs a generational replacement strategy. The proposed UDE algorithm is tested on the 28 benchmark problems provided for the CEC 2017 competition on constrained real parameter optimization. The experimental results demonstrate the efficacy of the presented algorithm in solving constrained real parameter optimization problems.

Decomposition is a well-known strategy in traditional multiobjective optimization. However, the decomposition strategy was not widely employed in evolutionary multiobjective optimization until Zhang and Li proposed multiobjective evolutionary algorithm based on decomposition (MOEA/D) in 2007. MOEA/D proposed by Zhang and Li decomposes a multiobjective optimization problem into a number of scalar optimization subproblems and optimizes them in a collaborative manner using an evolutionary algorithm (EA). Each subproblem is optimized by utilizing the information mainly from its several neighboring subproblems. Since the proposition of MOEA/D in 2007, decomposition-based MOEAs have attracted significant attention from the researchers. Investigations have been undertaken in several directions, including development of novel weight vector generation methods, use of new decomposition approaches, efficient allocation of computational resources, modifications in the reproduction operation, mating selection and replacement mechanism, hybridizing decomposition- and dominance-based approaches, etc. Furthermore, several attempts have been made at extending the decomposition-based framework to constrained multiobjective optimization, many-objective optimization, and incorporate the preference of decision makers. Additionally, there have been many attempts at application of decomposition-based MOEAs to solve complex real-world optimization problems. This paper presents a comprehensive survey of the decomposition-based MOEAs proposed in the last decade.

The increasing population of microgrids with various kinds of plug and play energy resources and rapidly varying demand in distribution systems are multiplying the complexity involved in overall system management. This paper proposes an agent-based transactive energy management framework with a comprehensive energy management system (CEMS) as a solution to address the aggregated complexity induced by microgrids in distribution systems. In this framework, microgrids sell or buy the energy in transactive market, which is an inter-microgrid auction based electricity market, to manage the excess supply or residual demand. CEMS follows a dual phase energy management strategy. In the first stage local auxiliary resources such as demand response and distributed energy storage systems of the microgrids are optimally integrated into system operation to level off the forecasted energy imbalances in microgrids. In the latter stage, the operating configuration of the local auxiliary resources is adjusted in real time along with transactive energy to address the imbalances leftover in the former phase and the forecast errors. The efficacy of the proposed framework and CEMS is verified on a IEEE distribution test feeder system with microgrids.

This paper is concerned with the leader-follower consensus of multiagent systems with wireless communications with the main objective of reducing the power consumption. First, by assuming that the sampling period jumps from one to another only from a given set, a new stochastic sampling approach is introduced to reduce the sampling frequency of each agent. Then, only 1-D of the sampled data is selected, and transmitted to its neighboring agents. Finally, each agent is scheduled to communicate with others intermittently. A unified Markovian system model is proposed to capture the above stochastic sampling, measurement selection scheme and intermittent transmission process, and such a novel protocol can significantly reduce the power consumption. Based on the Lyapunov stability theory and the Markovian jump system approach, the distributed consensus-based controller gain is obtained by solving an optimization problem. The advantage of the proposed consensus protocol is verified by two simulation examples. The simulation results explicitly show our result is more energy-efficient than that of existing one.

Autonomous robots such are unmanned aerial and unmanned ground vehicles are increasingly utilized in patrolling, surveillance, search and rescue and in missions that are hazardous for humans. Path-planning, path-generation and following a planned path successfully are fundamental requirements for autonomous operation of unmanned robots. Though the operational principle of aerial and ground robots are different, the algorithms for path-planning and path-following can be generalized. Vector Directed Path Generation and Tracking (VDPGT) proposed in this work is a platform independent path-generation and path-following algorithm. VDPGT is designed to dynamically adapt the shortest path to a destination. Simulation studies carried out on two ground robots (Turtlebot and Clearpath Husky), two aerial robots (AR Drone and Hector-quadrotor) and realtime experiments on Turtlebot and AR Drone demonstrate the platform independent nature of VDPGT.

The penetration of distribution generation (DGs) sources in the distribution grid is largely limited by various factors that involves feeder losses, power quality, stability and security of the grid. Different types of DGs have different impacts on the grid upon integration. The penetration of inverter-based DGs (IBDG) in the distribution networks are mainly limited due to their heavy impact on the system harmonic levels. Whereas, synchronous based DGs (SBDG) are mainly limited due to their high impact on fault current levels that affect the over-current relay coordination. Placement of existing power system components such as flexible alternating current transmission system (FACTS) devices that are generally used for voltage regulation and reactive power compensation can be suggested as an efficient solution for maximizing the penetration of DGs. This paper suggests a hybrid framework with a heuristic optimization algorithm for planning and allocation of DGs and FACTS devices considering collective grid constraints such as voltage harmonics constraints, power balance constraints, voltage regulation constraints and over-current relay constraints. The proposed approach is tested on the power distribution network of IEEE 30-bus system. Comparative analyses have been conducted to highlight the efficacy of the proposed optimization approach under various test case scenarios.

Smart Home is a modern home that allows residents to have high-level comfort with effective use of electricity. These objectives can be achieved by applying suitable and promising optimization algorithms and techniques. This paper presents a demand side management strategy which was integrated into the existing Home Energy Management System (HEMS). Home energy management system is a Multi-Agent System (MAS) based decentralized architecture proposed by the authors. This intelligent energy management system was developed on an IEEE FIPA (Foundation for Intelligent Physical Agents) compliant multi-agent platform. This enables agents to communicate, interact and negotiate with energy sources and devices of the smart home to provide the most efficient energy usage and minimize the cost of electricity bills. This also results some peak load shaving of the power distribution system of the smart home. Simulation studies show the potential of proposed multi-agent system technique together with the demand side management strategy to provide the optimum solution for smart home energy management.

This paper proposes an optimal charging scheduling strategy, which is based on an integrated grid-to-vehicle (G2V) and vehicle-to-grid (V2G) charging approach, for maximizing the penetration of plug-in electric vehicles (PEVs) within a workplace car park. The driving pattern of PEVs is modeled with statistical methods using probability density functions. Based on the developed driving pattern, a fuzzy inference system is designed to model the PEVs' energy requirement. A genetic algorithm (GA) with heuristic initialization is then utilized for performing the optimal charging scheduling of PEVs. The proposed strategy is implemented for charging of PEVs in a workplace car park and based on the evaluation of maximum possible PEV penetration, optimal location(s) are determined for the car park in the industrial and commercial laterals of a 38-node distribution system. The simulation results demonstrate that the optimal charging strategy can prove beneficial in: 1) minimizing the daily total cost incurred by the parking operator; 2) reducing the network peak load; 3) providing frequency regulation service; and 4) preventing the overloading of distribution transformer and distribution lines.

The non-dominated sorting genetic algorithm II (NSGA-II) is applied to the concurrent generation scheduling and load management of the isolated microgrid: the joint generation — voyage scheduling of an all-electric ship (AES). The AES uses integrated power generators and an energy storage system (ESS) to match its propulsion and service loads, thus forming an isolated microgrid. The propulsion load is the variable load that determines th ship voyage; hence, the schedulings of the generator, the propulsion load and the ESS have to be processed concurrently, for the optimal operation of the AES. NSGA-II outperforms the existing methods for the AES scheduling problem by modelling the schedulings in a one-stage optimisation. Another advantage of NSGA-II applied here is its multi-objective optimisation. Existing work on ship scheduling has hitherto set the minimisation of the AES operational cost as the single optimisation objective, with the reduction of the greenhouse gas (GHG) emission merely treated as a constraint of the optimisation. The proposed work emphasises the environmental concerns and makes GHG mitigation a separate objective, thus expanding the optimisation into multi-objective. The simulation demonstrate that optimising the ship voyage jointly with the generation scheduling results in the reduction of both operational cost and GHG emission, compared to the fixed-voyage generation scheduling. Integrating the ESS dispatch into the generation scheduling further enhances the benefits.

Multi-agent systems (MAS) have recently evolved as an important feature in the development of future smart grids, especially to provide the fast-responding self-healing ability to the grid. This paper presents a comparative analysis of centralized and decentralized MAS architecture for the problem of service restoration. Service restoration is formulated as a multi-objective optimization problem and solved by implementing DG islanding using both the centralized and decentralized MAS approach. Simulation studies are conducted on 38, 69, and 119 bus test distribution systems for both the MAS architectures. From the simulated results, the following advantages of the decentralized MAS architecture over the centralized MAS are observed: a) lower computation time to determine the restoration solution, b) robust to communication failure (not prone to single point failure), c) efficient restoration for multiple fault situations, d) lower communication cycles to discover the information.

In this paper, a multi-objective evolutionary algorithm based on decomposition (MOEA/D) is proposed to solve the unit commitment (UC) problem in presence of significant wind penetration as a multi-objective optimization problem considering cost, emission, and reliability as the multiple objectives. The uncertainties occurring due to thermal generator outage, load forecast error, and wind forecast error are incorporated using expected energy not served (EENS) reliability index and EENS cost is used to reflect the reliability objective. Since, UC is a mixed-integer optimization problem, a hybrid strategy is integrated within the framework of MOEA/D such that genetic algorithm (GA) evolves the binary variables while differential evolution (DE) evolves the continuous variables. The performance of the proposed algorithm is investigated on a 20 unit test system. To improve the performance of the algorithm in terms of distribution of solutions obtained, an external archive strategy based on ε-dominance principle is implemented. The simulation results demonstrate that the proposed algorithm can efficiently obtain a well-distributed set of trade-off solutions on the multi-objective wind-thermal UC problem.

This paper proposes a new method for global optimization through randomized group search in contracting regions. For each iteration, a population is randomly produced within the search region, where the population size is chosen to ensure that the empirical optimum is an estimate of the true optimum within a predefined accuracy with a certain confidence. Fitness values are evaluated at the samples in the population. A very small subset of them with top-ranking fitness values are selected as good points. Neighborhoods of these good points are used to form a new and smaller search region, in which a new population is generated. It is easy to implement the algorithm. Extensive simulation on benchmark problems shows that the proposed method is fast and reasonably accurate.

An emerging scenario of using Electric Vehicles to provide reactive power support arises as the population of Electric Vehicles (EVs) increases significantly. In view of this emerging scenario, a new application of Particle Swarm Optimization is proposed in this paper, which is adopting Particle Swarm Optimization (PSO) in Vehicle-to-Grid (V2G) based reactive power dispatch. Rather than utilizing reactive power support from central generators, the proposed method aims at increasing reactive power support from EVs, which are distributed. Case study on a modified IEEE 9-bus system is done and results show that the proposed method can bring convincing benefits on voltage stability and power loss in the overall power system.

The Dual Active Bridge (DAB) topology is ideally suited for high-power dc/dc conversion especially for energy storage applications. Due to the low current ripple and higher conversion requirements, Current-Fed DAB (CFDAB) with secondary-side modulation is implemented to avoid the voltage spike across parasitic capacitance of Low Voltage (LV) switches. This paper details Phase Shift Modulation (PSM) and Phase Shift plus High Voltage Duty Modulation (PSDM) implemented to transfer the power and avoid voltage spike. A comparison between the two in terms of the power transfer, soft-switching conditions and transformer RMS and circulating currents are detailed. Also, to improve the efficiency of the converter at low load conditions, an Extended Phase Shift Modulation (EPSM) is proposed. The zero state in High Voltage (HV) transformer voltage is introduced by the phase shift between the cross-connected legs in HV. The phase shift between the legs reduces to zero at heavy loads and thus, the modulation shifts to PSM smoothly. The zero state reduces the peak current which thus, reduces the back flow power. The above Modulation analyses power transfer and transformer currents within the safe operation region of the converter. Simulation results are shown to validate the theoretical claims for full load and partial load conditions for PSM and EPSM.

Energy Management Systems have become an imperative aspect of smart grids, owing to the enormous challenges imposed due to real-time pricing, distributed generation and integration of intermittent renewables. Due to the uncertainty associated with renewable sources and prominent fluctuations in the load demand, it is extremely important to maintain the overall energy balance in such grids. In this paper, the distributed energy management is achieved using a Multi-agent System which provides a flexible and reliable solution to control and manage smart grids. Adaptive fuzzy systems are designed to instill intelligent decision making capability in the agents of multi-agent system. When renewable sources are inadequate, the sustainability of the system is not guaranteed and multi-agent system is capable of deciding the mode of operation such that the system reliability and performance is not compromised. The proposed algorithm maintains power balance in the system and also sustains desired values for the State of Charge of storage units in order to guarantee extended battery life. The Energy management system also implements a cost optimization algorithm based on the Particle Swarm Optimization technique, to minimize operating costs and maximize profits earned by the grid. The proposed energy management algorithm is tested and validated on a practical test system which inherits most of the features of a small-scale smart grid.

The electric power grid is spatially and temporally complex, nonconvex, nonlinear, and nonstationary system with uncertainties at many levels. The integration of renewable sources of energy, such as wind and solar farms, energy storage, and plug-in hybrid electric vehicles, further adds complexity and challenges to efficient, reliable, and safe operation of electric power grids. A smart grid is aimed at improving power system’s reliability, security, sustainability, efficiency, and flexibility through distributed and coordinated intelligence at all levels of the electric power grid—generation, transmission, and distribution. The challenges faced in evolving a smart grid include variable and uncertain generation (such as wind and solar), stochastic load profiles and power flows, cyber attacks (unintentional and malicious), communication latencies and data loss, and so on.

Electricity markets in countries around the world are being restructured in pursuit of economic efficiency through competition. However, unpredictability in electricity prices and previous occurrences of market failure have indicated a need to better understand the complex interactions between the various market participants as well as to design market rules that maximizes efficiency and security. In this paper, the techniques of Agent-based Computational Economics (ACE) are employed to simulate the behavior of the GenCo participants in the National Electricity Market of Singapore (NEMS). The use of Reinforcement Learning (RL) in the agent-based modeling will be a more realistic representation of the GenCo and will bring about more accurate electricity market simulation outcomes. Actor-Critics constitute an important aspect of RL and in this paper we propose an adaptive actor-critic mapping using Particle Swarm Optimization (PSO). A simulation platform is built with the proposed model for Genco's learning and is tested in the conditions of varying vesting contract levels. Simulation results indicate that the proposed learning algorithm is able to procure higher GenCo revenue when benchmarked with existing learning algorithms.

This paper addresses day-ahead unit commitment with significant wind penetration as a multi-objective optimization problem in an uncertain environment considering system operation cost and reliability as the multiple conflicting objectives. The uncertainties occurring due to thermal unit outage, load forecast error and wind forecast error are efficiently incorporated using expected energy not served (EENS) reliability index while EENS cost is used to reflect the reliability objective. A multi-objective genetic algorithm is proposed to solve the aforementioned scheduling problem. The algorithm is implemented on a 20 unit test system and the effect of wind penetration level, value of lost load and load forecast uncertainty is analyzed. It is demonstrated and validated through simulation studies that optimum system spinning reserve is the amount for which the sum of system operation cost and expected energy not served cost i.e., total cost is minimum. Amongst the trade-off optimal solutions obtained, a single optimum solution is highlighted which can be most important to system operators.

The high penetration of Distributed Generators (DGs) increases the need for monitoring and protection of the distribution system. The stochastic nature of the DGs may result in varying fault currents seen by the conventional over-current protection relays and thereby disturb the coordination of the relays. This necessitates an effective numerical relay that can capture the changes in the varying nature of DGs and take effective decisions according to the changing network conditions. Hence, an adaptive fuzzy relay, comprising of a fuzzy inference module and a neural network learning module, has been developed for deciding the optimal protection settings in the numerical relay corresponding to the changes in the network scenarios. A systematic comparison of the proposed adaptive fuzzy relay with conventional relay has been presented on a standard IEEE-test distribution system. The simulation results verify that the adaptive fuzzy relay is able to achieve the desired protection settings using a closed-loop approach.

In this paper, a multiobjective evolutionary algorithm based on decomposition (MOEA/D) is proposed to solve the unit commitment (UC) problem as a multiobjective optimization problem (MOP) considering minimizing cost and emission as the multiple objectives. Since UC problem is a mixed-integer optimization problem, a hybrid strategy is integrated within the framework of MOEA/D such that genetic algorithm (GA) evolves the binary variables, while differential evolution (DE) evolves the continuous variables. Further, a novel nonuniform weight-vector distribution (NUWD) strategy is proposed and an ensemble algorithm based on combination of MOEA/D with uniform weight-vector distribution (UWD) and NUWD strategy is implemented to enhance the performance of the presented algorithm. Extensive case studies are presented on different test systems and the effectiveness of the hybrid strategy, the NUWD strategy, and the ensemble algorithm is verified through stringent simulated results. Further, exhaustive benchmarking against the algorithm proposed in the literature is presented to demonstrate the superiority of the proposed algorithm.

This brief proposes an efficient technique for the construction of optimized prediction intervals (PIs) by using the bootstrap technique. The method employs an innovative PI-based cost function in the training of neural networks (NNs) used for estimation of the target variance in the bootstrap method. An optimization algorithm is developed for minimization of the cost function and adjustment of NN parameters. The performance of the optimized bootstrap method is examined for seven synthetic and real-world case studies. It is shown that application of the proposed method improves the quality of constructed PIs by more than 28% over the existing technique, leading to narrower PIs with a coverage probability greater than the nominal confidence level.

Penetration of renewable energy resources, such as wind and solar power, into power systems significantly increases the uncertainties on system operation, stability, and reliability in smart grids. In this paper, the nonparametric neural network-based prediction intervals (PIs) are implemented for forecast uncertainty quantification. Instead of a single level PI, wind power forecast uncertainties are represented in a list of PIs. These PIs are then decomposed into quantiles of wind power. A new scenario generation method is proposed to handle wind power forecast uncertainties. For each hour, an empirical cumulative distribution function (ECDF) is fitted to these quantile points. The Monte Carlo simulation method is used to generate scenarios from the ECDF. Then the wind power scenarios are incorporated into a stochastic security-constrained unit commitment (SCUC) model. The heuristic genetic algorithm is utilized to solve the stochastic SCUC problem. Five deterministic and four stochastic case studies incorporated with interval forecasts of wind power are implemented. The results of these cases are presented and discussed together. Generation costs, and the scheduled and real-time economic dispatch reserves of different unit commitment strategies are compared. The experimental results show that the stochastic model is more robust than deterministic ones and, thus, decreases the risk in system operations of smart grids.

This paper proposes a high frequency soft switched high voltage gain dc/dc converter for DC microgrid application. The proposed converter employs a half-bridge resonant boost converter at input and a voltage quadruple circuit at output. Resonant boost converter is operated at frequency of 150 kHz to gain advantage of low output voltage ripple and reduced magnetics. Zero voltage turn-on is achieved for all switches. Zero current turn-on and turn-off is achieved for all diodes. High frequency film capacitors increase life time of the converter. Voltage stress across switches is less and is clamped naturally without external snubber circuit. Experimental converter rated at 300 W has been designed, and tested to verify the analysis, design and demonstrate the performance of the proposed converter.

Electrical power systems are evolving from today's centralized bulk systems to more decentralized systems. Penetrations of renewable energies, such as wind and solar power, significantly increase the level of uncertainty in power systems. Accurate load forecasting becomes more complex, yet more important for management of power systems. Traditional methods for generating point forecasts of load demands cannot properly handle uncertainties in system operations. To quantify potential uncertainties associated with forecasts, this paper implements a neural network (NN)-based method for the construction of prediction intervals (PIs). A newly introduced method, called lower upper bound estimation (LUBE), is applied and extended to develop PIs using NN models. A new problem formulation is proposed, which translates the primary multiobjective problem into a constrained single-objective problem. Compared with the cost function, this new formulation is closer to the primary problem and has fewer parameters. Particle swarm optimization (PSO) integrated with the mutation operator is used to solve the problem. Electrical demands from Singapore and New South Wales (Australia), as well as wind power generation from Capital Wind Farm, are used to validate the PSO-based LUBE method. Comparative results show that the proposed method can construct higher quality PIs for load and wind power generation forecasts in a short time.

The electricity market has provided a complex economic environment, and consequently has increased the requirement for advancement of learning methods. In the agent-based modeling and simulation framework of this economic system, the generation company's decision-making is modeled using reinforcement learning. Existing learning methods that model the generation company's strategic bidding behavior are not adapted to the non-stationary and non-Markovian environment involving multidimensional and continuous state and action spaces. This paper proposes a reinforcement learning method to overcome these limitations. The proposed method discovers the input space structure through the self-organizing map, exploits learned experience through Roth-Erev reinforcement learning and explores through the actor critic map. Simulation results from experiments show that the proposed method outperforms Simulated Annealing Q-Learning and Variant Roth-Erev reinforcement learning. The proposed method is a step towards more realistic agent learning in Agent-based Computational Economics.

Optimal selection and sizing of Distributed Energy Resources (DER) is an important research problem for the advancement of distributed power systems. This paper presents detail studies on optimal sizing of DER for integrated microgrids using Evolutionary Strategy (ES). Integrated microgrid is an innovative architecture in distributed power systems, in which several microgrids are interconnected with each other for superior control and management of the distributed power systems. Right coordination among DER in microgrids, and proper harmony among the microgrids and the main distribution grid are critical challenges. Types of DER and capacities of them are needed to optimize such that proposed integrated microgrid provides reliable supply of energy at cheap cost. In this research, the problem is formulated as a nonlinear mixed-integer minimization problem which minimizes capital and annual operational cost of DER subject to a variety of system and unit constraints. Evolutionary strategy was developed for solving the minimization problem. The proposed methodology was used to design integrated microgrids for A*Star IEDS (Intelligent Energy Distribution System) project. The design results have shown that the proposed methodology provides excellent convergence and feasible optimum solution.

The short-term unit commitment problem is traditionally solved as a single-objective optimization problem with system operation cost as the only objective. This paper presents multi-objectivization of the short-term unit commitment problem in uncertain environment by considering reliability as an additional objective along with the economic objective. The uncertainties occurring due to unit outage and load forecast error are incorporated using loss of load probability (LOLP) and expected unserved energy (EUE) reliability indices. The multi-objectivized unit commitment problem in uncertain environment is solved using our earlier proposed multi-objective evolutionary algorithm [1]. Simulations are performed on a test system of 26 thermal generating units and the results obtained are benchmarked against the study [2] where the unit commitment problem was solved as a reliability-constrained single-objective optimization problem. The simulation results demonstrate that the proposed multi-objectivized approach can find solutions with considerably lower cost than those obtained in the benchmark. Further, the efficiency and consistency of the proposed algorithm for multi-objectivized unit commitment problem is demonstrated by quantitative performance assessment using hypervolume indicator.

Prediction intervals (PIs) are excellent tools for quantification of uncertainties associated with point forecasts and predictions. This paper adopts and develops the lower upper bound estimation (LUBE) method for construction of PIs using neural network (NN) models. This method is fast and simple and does not require calculation of heavy matrices, as required by traditional methods. Besides, it makes no assumption about the data distribution. A new width-based index is proposed to quantitatively check how much PIs are informative. Using this measure and the coverage probability of PIs, a multi-objective optimization problem is formulated to train NN models in the LUBE method. The optimization problem is then transformed into a training problem through definition of a PI-based cost function. Particle swarm optimization (PSO) with the mutation operator is used to minimize the cost function. Experiments with synthetic and real-world case studies indicate that the proposed PSO-based LUBE method can construct higher quality PIs in a simpler and faster manner.

Deregulated power industries increase the efficiency of electricity production and distribution, and offer higher quality, secure, and more reliable electricity at low prices. In a deregulated environment, utilities are not required to meet the total load demand. Generation companies (GENCOs) schedule the generators that produce less than the predicted load demand and reserve, but aim to deliver maximum profits. The scheduling of generators depends on the market price. More number of generating units are committed when the market price is higher. When more number of generating units are brought in the deregulated market, more profit can be achieved by producing higher amount of power. This paper present a hybrid algorithm to solve a profit based unit commitment problem in a deregulated environment. The proposed algorithm has been developed from generation company's point of view. It maximizes the profit of the generation company in the deregulated power and reserve markets. A hybrid methodology between Lagrangian Relaxation and Generic Algorithm (LRGA) is used to solve generation scheduling in a day-ahead competitive electricity market. The results obtained are quite encouraging and useful in deregulated market optimization.

Deregulation in the electricity market offers freedom to the generator companies (GENCOs) to schedule their generators in order to maximize their profit without actually satisfying the load and the reserve requirements. Various techniques have been developed for solving the profit based unit commitment (PBUC) problem. Among them, the multi-agent approach is different where each generator unit is referred to as an intelligent agent. In this paper, we develop a new multi-agent approach for PBUC problem in which the rule based intelligence is provided to the independent system operator (ISO) agent. Intelligence of generator agents (GenAgents) is limited to maximize their profit for the given demand and reserve using real-parametric genetic algorithm (GA) and share the results with ISO agent. In this approach, ISO agent commits the maximum profit generating GenAgents for every hour while satisfying the up/down time constraints. ISO agent also asks other GenAgents to calculate their profit for the remaining demand and reserve. The simulation results of 10 units problem for two payment methods are shown and compared with other techniques.

The deregulation of the electrical power industries has opened many opportunities to power buyers, once price takers of a monopolistic economy, to look forward to a free market economy with market forces determining the market clearing prices and quantities. However, the strong influence of technical and physical constraints of the network may result in economic decisions that adversely affect the interests of the consumers. Compared to the monopolistic economy of yesteryears, power buyers may actually be able to influence the market by cooperating with other power buyers in the network. This paper presents a co-evolutionary algorithm for evolving individual and cooperative strategies of electricity buyers in a power market. The algorithm focuses on how the buyers choose their bidding strategies through learning to maximize the profits in different scenarios of playing individually or cooperatively. The results show that it is of great benefit to cooperate but the free rider problem may arise when an individual buyer gains more profit due to the cooperative effort of the others.

This paper presents a multi-objective evolutionary algorithm to solve the day-ahead thermal generation scheduling problem. The objective functions considered to model the scheduling problem are: 1) minimizing the system operation cost and 2) minimizing the emission cost. In the proposed algorithm, the chromosome is formulated as a binary unit commitment matrix (UCM) which stores the generator on/off states and a real power matrix (RPM) which stores the corresponding power dispatch. Problem specific binary genetic operators act on the binary UCM and real genetic operators act on the RPM to effectively explore the large binary and real search spaces separately. Heuristics are used in the initial population by seeding the random population with two Priority list (PL) based solutions for faster convergence. Intelligent repair operator based on PL is designed to repair the solutions for load demand equality constraint violation. The ranking, selection and elitism methods are borrowed from NSGA-II. The proposed algorithm is applied to a large scale 60 generating unit power system and the simulation results are presented and compared with our earlier algorithm [26]. The presented algorithm is found to outperform our earlier algorithm in terms of both convergence and spread in the final Pareto-optimal front.

This paper presents a new solution to thermal unit commitment (UC) problem based on a modified hybrid particle swarm optimization (MHPSO). Hybrid real and binary PSO is coupled with the proposed heuristic based constraint satisfaction strategy that makes the solutions/particles feasible for PSO. The velocity equation of particle is also modified to prevent particle stagnation. Unit commitment priority is used to enhance the performance of binary PSO. The proposed algorithm is tested for 10, 20, 40 and 60 unit systems and the results are reported for 10 different runs. Statistical results and their comparison show a good performance of MHPSO over other existing optimization methods.

The bootstrap method is one of the most widely used methods in literature for construction of confidence and prediction intervals. This paper proposes a new method for improving the quality of bootstrap-based prediction intervals. The core of the proposed method is a prediction interval-based cost function, which is used for training neural networks. A simulated annealing method is applied for minimization of the cost function and neural network parameter adjustment. The developed neural networks are then used for estimation of the target variance. Through experiments and simulations it is shown that the proposed method can be used to construct better quality bootstrap-based prediction intervals. The optimized prediction intervals have narrower widths with a greater coverage probability compared to traditional bootstrap-based prediction intervals.

Most of the real world physical systems have critical thresholds, also known as tipping points, at which the system abruptly shifts its state from one to another. From dynamical system's perspective, bifurcation is the phenomenon responsible for these critical transitions in the system. There are various directions which can be adopted to study this bifurcation problem in an attempt to predict this phenomenon. The focus of this paper is classical bifurcation theory based approach incorporating multiple scale dynamics which is able to give analysis of bifurcations responsible for critical transitions in electricity price time series system. Fitz-Hugh Nagumo (FHN), which is a classical example exhibiting slow-fast scale dynamics is studied and later on hybridized with nonlinear neural networks to model this time series in various markets. Encouraging results allow us to look into this approach in future.

Evolving artificial neural networks has attracted much attention among researchers recently, especially in the fields where plenty of data exist but explanatory theories and models are lacking or based upon too many simplifying assumptions. Financial time series forecasting is one of them. A hybrid model is used to forecast the hourly electricity price from the California Power Exchange. A collaborative approach is adopted to combine ANN and evolutionary algorithm. The main contributions of this thesis include: Investigated the effect of changing values of several important parameters on the performance of the model, and selected the best combination of these parameters; good forecasting results have been obtained with the implemented hybrid model when the best combination of parameters is used. The lowest MAPE through a single run is 5.28134%. And the lowest averaged MAPE over 10 runs is 6.088%, over 30 runs is 6.786%; through the investigation of the parameter period, it is found that by including “future values” of the homogenous moments of the instant being forecasted into the input vector, forecasting accuracy is greatly enhanced. A comparison of results with other works reported in the literature shows that the proposed model gives superior performance on the same data set.

Preference-based techniques in multi-objective evolutionary algorithms (MOEA) are gaining importance. This paper presents a method of representing, eliciting and integrating decision making preference expressed as a set of imprecise goal vectors into a MOEA with steady-state replacement. The specification of a precise goal vector without extensive knowledge of problem behavior often leads to undesirable results. The approach proposed in this paper facilitates the linguistic specification of goal vectors relative to extreme, non-dominated solutions (i.e. the goal is specified as ”Very Small”, ”Small”, ”Medium”, ”Large”, and ”Very Large”) with three degrees of imprecision as desired by the decision maker. The degree of imprecision corresponds to the density of solutions desired within the target subset. Empirical investigations of the proposed method yield promising results.

In this paper, two TS-type fuzzy logic controllers (FLC) in direct and incremental forms are constructed for freeway local ramp metering tasks. The values of consequent part of fuzzy rules are optimized with a particle swarm optimization algorithm (PSO). The optimization process under PSO is carried out on PARAMICS microscopic traffic simulation platform. FLC methods and the traditional ALINEA control method are examined and compared on the traffic density performance. Simulation results on PARAMICS show the applicability and efficiency of the proposed FLC.

In this paper a general methodology is proposed for development of spiking neural networks (SNN) as a time series modeling task. A continuous firing temporal encoding scheme is employed in the developed model for efficient handling of temporal correlations in high dimensional chaotic time series. The universal nonlinear function approximation property and unique ability of temporally encoded SNN is particularly advantageous in complex dynamics scenario. Rich dynamics of spiking neural networks are exploited for forecasting in electricity price time series system. The temporal encoding scheme proposed particularly for time series applications produced interesting results which encourage further research in this direction.

This paper describes the use of decision maker preferences in terms of the relative importance of objectives in evolutionary multiobjective optimization. A mathematical model of the relative importance of objectives and an elicitation algorithm are proposed, and three methods of incorporating explicated preference information are described and applied to standard test problems in an empirical study. The axiomatic model proposed here formalizes the notion of relative importance of objectives as a partial order that supports strict preference, equality of importance, and incomparability between objective pairs. Unlike most approaches, the proposed model does not encode relative importance as a set of real-valued parameters. Instead, the approach provides a functional correspondence between a coherent overall preference with a subset of the Pareto-optimal front. An elicitation algorithm is also provided to assist a human decision maker in constructing a coherent overall preference. Besides elicitation ofapriori preference, an interactive facility is also furnished to enable modification of overall preference while the search progresses. Three techniques of integrating explicated preference information into the well-known Non-dominated Sorting Genetic Algorithm (NSGA)-II are also described and validated in a set of empirical investigation. The approach allows a focus on a subset of the Pareto-front. Validations on test problems demonstrate that the preference-based algorithm gained better convergence as the dimensionality of the problems increased.

Rapid urbanization and the growing demand for faster transportation has led to heavy congestion in road traffic networks, necessitating the need for traffic-responsive intelligent signal control systems. The developed signal control system must be capable of determining the green time that minimizes the network-wide travel time delay based on limited information of the environment. This paper adopts a distributed multiagent-based approach to develop a traffic-responsive signal control system, i.e., the geometric fuzzy multiagent system (GFMAS), which is based on a geometric type-2 fuzzy inference system. GFMAS is capable of handling the various levels of uncertainty found in the inputs and rule base of the traffic signal controller. Simulation models of the agents designed in PARAMICS were tested on virtual road network replicating a section of the central business district in Singapore. A comprehensive analysis and comparison was performed against the existing traffic-control algorithms green link determining (GLIDE) and hierarchical multiagent system (HMS). The proposed GFMAS signal control outperformed both the benchmarks when tested for typical traffic-flow scenarios. Further tests show the superior performance of the proposed GFMAS in handling unplanned and planned incidents and obstructions. The promising results demonstrate the efficiency of the proposed multiagent architecture and scope for future development.

Rapid advances made in vehicle technology and increased level of urbanization have caused an exponential increase in road traffic congestion levels. This has necessitated the implementation of intelligent traffic responsive signal controllers capable of maintaining the saturation levels in each link thereby reducing congestion and increasing utilization of existing infrastructure. This paper presents one such distributed multi-agent architecture based on weighted type-2 fuzzy inference engine for the urban traffic signal control. Agents have been programmed in PARAMICS microscopic traffic simulator and tested on a simulated section of Central Business District in Singapore with twenty five interconnected intersections. A comparative analysis of the proposed architecture with the existing traffic signal controller HMS - Hierarchical multi-agent system, was performed for two different traffic scenarios. The results clearly indicates better performance of the proposed agent architecture over the benchmark controller and offers scope for improvement in the future.

The optimal solutions of a multiobjective optimization problem correspond to a nondominated front that is characterized by a tradeoff between objectives. A knee region in this Pareto-optimal front, which is visually a convex bulge in the front, is important to decision makers in practical contexts, as it often constitutes the optimum in tradeoff, i.e. substitution of a given Pareto-optimal solution with another solution on the knee region yields the largest improvement per unit degradation. This paper presents a selection scheme that enables a multiobjective evolutionary algorithm (MOEA) to obtain a nondominated set with controllable concentration around existing knee regions of the Pareto front. The preference- based focus is achieved by optimizing a set of linear weighted sums of the original objectives, and control of the extent of the focus is attained by careful selection of the weight set based on a user-specified parameter. The fitness scheme could be easily adopted in any Pareto-based MOEA with little additional computational cost. Simulations on various two- and three- objective test problems demonstrate the ability of the proposed method to guide the population toward existing knee regions on the Pareto front. Comparison with general-purpose Pareto based MOEA demonstrates that convergence on the Pareto front is not compromised by imposing the preference-based bias. The performance of the method in terms of an additional performance metric introduced to measure the accuracy of resulting convergence on the desired regions validates the efficacy of the method.

The ease with which a human expert driver performs the complex tasks involved in parallel-parking a non-holonomic vehicle motivates the mimicry of an human driving behavior in automation of the task. This paper presents such an algorithm to achieve automated parallel parking in tight spaces. Unlike other approaches rooted in neural networks and/or fuzzy logic, the proposed algorithm performs maneuvers closely modeled after human driving instructions. Stevenspsila power law is employed in modeling perceived physical quantities on which the instructions operate while the uncertainty inherent in the natural language formulation is represented by Gaussian distribution. The algorithm consists of five stages: position alignment in preparation for the backward S-turn, the first half of the S-turn, position alignment for the second part of the S-turn, the second part of the S-turn and longitudinal adjustment. Negotiation of available parking space in the second part of the S-turn, arguably the most difficult part, is performed with the help of a rule base documenting the relation between steering angle, vehicle orientation and distance traversed. To achieve parking accuracy and avoid collision in the maneuver, the appropriate steering angle must be employed. This angle is approximated from the most suitable rule, which identification is essentially a multi-objective problem addressed here by a Multi-Objective Evolutionary Algorithm. Computer simulations demonstrate the success of the approach.

Coordination is one of the key components in distributed multi-agent systems. Establishing a coordination scheme with minimum communication requirements and robustness to communication failure is a difficult task. A new multi-agent architecture based on type-2 fuzzy decision making is proposed here for achieving coordination with minimum communication. The decision making module has been designed to achieve the coordination between agents by calculating the weight of the input to be used for deciding on the action plans in a dynamic manner. The effectiveness of the coordination scheme proposed was tested by applying it to a complex, non-linear and stochastic application of the traffic signal control. The size of the network chosen also serves to show the scalability of the agent architecture. The results obtained were compared with adaptive systems, fixed coordination schemes and no coordination schemes. Considerable improvement in the time delay was achieved while using the dynamic coordination scheme proposed.

Road Traffic congestion can occur anywhere from normal city roads, freeways to even highways. Traffic congestion can also be accentuated by incidents like terrorist attacks, accidents and breakdowns. This paper summarizes the use of various evolutionary techniques for traffic management and congestion avoidance in Intelligent Transportation Systems. Evolutionary algorithms with their inherent strength as optimization techniques are good candidates for solutions to road traffic management and congestion avoidance problems. A number of approaches involving the use of Genetic algorithms, Learning Classifier Systems and Genetic programming have been discussed for solutions to different problems in this domain. This paper proposes a multi-agent based real-time centralized evolutionary optimization technique for urban traffic management in the area of traffic signal control. This scheme uses evolutionary strategy for the control of traffic signal. The total vehicle mean delay in a six junction network was reduced by using evolutionary strategy. In order to achieve this the green signal time was optimized in an online manner. Comparison with a fixed time based traffic controller has been made and was found to produce better results.

Real-world applications are bound to have certain level of uncertainty inherent in them. Among this noise is one of the most predominant factors affecting the optimization process whether it is conventional or evolutionary techniques. The evolutionary optimization techniques are found to be inherently stronger and robust to noisy environments but they are robust for lower noise levels, higher noise requires corrections to be made to the algorithm. This paper attempts to provide a comprehensive overview of the different correction methods used for optimizing noisy objective functions or fitness functions that creates uncertain environment and also provide with an brief overview of the other issues involved while using evolutionary computational methods for optimizing applications in uncertain environment.

This paper considers a dynamic route planning problem (DRPP) involving the optimization of a route for a single vehicle traveling between a given source and given destination. Although route planning has been widely studied, most of the available applications are primarily targeted at finding the shortest path (SP) routes, which is insufficient for dynamic route planning in real life scenario. For example, the travel time for the SP may not correspond to the overall shortest time (ST) route due to varying road conditions. In this paper, the proposed Hybrid Evolutionary Algorithm for solving the Dynamic Route Planning Problem (HEADRPP) is believed to be capable of solving this problem. The proposed HEADRPP comprises a Fuzzy Logic Implementation (FLI) and a Graph Partitioning Algorithm (GPA) incorporated into a Genetic Algorithm (GA) core, and offers both optimized SP and ST routes to the user. In this paper, the proposed HEADRPP is successfully tested on a 138 node network extracted from the Singapore Map, and its performance on SP optimization is compared with a pure GA and an ant based algorithm. Overall the performance of the proposed HEADRPP is shown to be robust to the dynamic nature of the DRPP.

Finding the shortest path from a given source to a given destination is a well known and widely applicable problem. Most of the work done in the area have used static route planning algorithms such as A*, Dijkstra’s, Bellman-Ford algorithm etc. Although these algorithms are said to be optimum, they are not capable of dealing with certain real life scenarios. For example, most of these single objective optimizations fails to find the equally good solutions when there is more than one optimum (shortest distance path, least congested path). We believe that the Genetic Algorithm (GA) based route planning algorithm proposed in this paper has the ability to tackle the above problems. In this paper, the proposed GA based route planning algorithm is successfully tested on the entire Singapore map with more than 10,000 nodes. Performance of the proposed GA is compared with an ant based path planning algorithm. Simulation results demonstrate the effectiveness of the proposed algorithm over ant based algorithm. Moreover, the proposed GA may be used as a basis for developing an intelligent route planning system.

This paper summarizes the collective work done in the application of evolutionary computation for financial time series forecasting. These are mainly stock market indices and foreign exchange rate prediction. The time series corresponding to these indices is a non-linear dynamic stochastic system different from other static patterns which are independent of time. Evolutionary techniques have capabilities of efficient search space exploration with population models corresponding to the problem. Their ability to capture the non linear dependencies among the system variables has invited economic analysts towards their use in the field of financial time series prediction. In this paper, previous research done in the application of evolutionary techniques for economic time series prediction and resolving the issues involved has been presented.

This paper presents the application of two co- evolutionary algorithms for evolving buyers' bidding strategies in a restructured pool-type electrical power market. A "greedy" algorithm which always aims to get higher power and pay less Ideational marginal price, as well as a "demand- driven" algorithm which aims to follow closely the individual demand, have been analyzed and implemented in simulations under different market scenarios. The two distinctive algorithms were compared against each other in a simulated power market of a reasonably large scale with 7 buyers and 20 sellers in an IEEE 14 bus network. The PowerWorldreg simulator has been used as a tool to ensure that the system validity and various constraints have been met. The simulation results suggest that a "demand-driven" co-evolutionary algorithm is more effective as it does not only help buyers to save cost when supply in the market is sufficient, but also enables them to outbid their opponents easily during tougher situations, such as when supply is in great shortage.

This paper introduces a dual layered particle swarm optimization algorithm (DLPSO), an evolutionary algorithm proposed to design an artificial neural network (ANN). The algorithm evolves the architecture of the ANN and trains its weights simultaneously. Different from the other techniques previously used, the proposed algorithm evolves the architecture along with the weights in two different layers. Tested on a non-linear system, typically a boost converter, the DLPSO evolves an optimal ANN controller to produce more efficient and robust results than the conventional control techniques used. The performance of the DLPSO based ANN controller is compared to that of a conventional PI controller at different operating points of the non-linear system. The tests show that the evolved controller performs equal to or better than the conventional techniques in terms of overshoot voltages and settling times for small and large signal input transients. Also, a comparison between the applicability of a PSO and a real-valued genetic algorithm for the training of weights is presented which shows that the PSO is faster and more efficient as a learning algorithm. Moreover, the proposed approach fully automates the neural network generation process, thus removing the need for time consuming manual design.

This paper proposes a reduced multivariate polynomial based pattern classifier using a three-layer neural network with linear regularized least square algorithm as an adjunct operation, for freeway incident detection. Freeway incident detection can be seen as a two class pattern classification problem where the rate of convergence is a major concern besides accurate classification. The reduced multivariate polynomial based model is particularly suitable for simple classification problems with small number of features and with large number of patterns available. Freeway incident detection is one such class of problem. Smaller number of terms in the reduced model compared to original full multivariate polynomial model results in small network size and increased speed of convergence, thus making it useful for freeway incident detection. SVD based and gradient descent based least square estimators were used separately and encouraging results were obtained compared to other classification strategies used for freeway incident detection allowing for further work on the use of this model with improvement in the algorithm.

This paper presents a review of preference incorporation in Multi-Objective Evolutionary Algorithms (MOEA). The incorporation of preference in Evolutionary Multi-objective Optimization (EMO) promotes better decisionmaking. Introducing preference in MOEAs increases the specificity of selection, leading to solutions which are of higher relevance to the Decision Maker(s). When many objectives are involved, a MOEA based on pure Pareto-optimality criterion may not achieve meaningful search. The incorporation of preference addresses this concern. The incorporation of preference is difficult because of uncertainties arising from lack of prior problem knowledge and fuzziness of human preference. Further, decision making is a complex and ill-defined process which at times could not be mathematically characterized. These concerns must be addressed in the incorporation of preference.

A knee region on the Pareto-optimal front of a multi-objective optimization problem consists of solutions with the maximum marginal rates of return, i.e. solutions for which an improvement on one objective is accompanied by a severe degradation in another. The trade-off characteristic renders such solutions of particular interest in practical applications. This paper presents a multi-objective evolutionary algorithm focused on the knee regions. The algorithm facilitates better decision making in contexts where high marginal rates of return are desirable by providing the decision makers with a high concentration of solutions on the knee regions of the Pareto-front approximation. The proposed approach computes a transformation of the original objectives based on weighted-sum functions. The transformed functions identify niches which correspond to knee regions in the objective space. The extent and density of coverage of the knee regions are controllable by the niche strength and pool size parameters. Although based on weighted-sums, the algorithm is capable of finding solutions in the non-convex regions of the Pareto-front. The application of the algorithm on test problems with multiple knee regions and skew on the Pareto-optimal front produces promising results.

This paper presents an evolutionary algorithm to generate cooperative strategies for individual buyers in a competitive power market. The paper explores how buyers can lower their costs by using an evolutionary algorithm that evolves their group sizes and memberships. The evolutionary process uncovers interesting agent behaviors and strategies for collaboration. The developed agent-based model uses PowerWorld simulator to incorporate the traditional physical system characteristics and constraints while evaluating individual agent’s behavior, actions and reactions on market dynamics. Simulation results on IEEE 14-bus system show that the evolutionary approach evolves mutually beneficial strategies that enhance the buyer’s profitability. The buyers learn to achieve substantial cost savings by forming groups and adjusting their demand curves, without sacrificing much in desired power consumption.

A knee region on the Pareto-optimal front of a multi-objective optimization problem consists of solutions with the maximum marginal rates of return, i.e. solutions for which an improvement on one objective is accompanied by a severe degradation in another. The trade-off characteristic renders such solutions of particular interest in practical applications. This paper presents a multi-objective evolutionary algorithm focused on the knee regions. The algorithm facilitates better decision making in contexts where high marginal rates of return are desirable for Decision Makers. The proposed approach computes a transformation of the original objectives based on weighted-sum functions. The transformed functions identify niches which correspond to knee regions in the objective space. The extent and density of coverage of the knee regions are controllable by the niche strength and pool size parameters. Although based on weighted-sums, the algorithm is capable of finding solutions in the non-convex regions of the Pareto-front.

The generic Multi-objective Evolutionary Algorithm (MOEA) aims to produce Pareto-front approximations with good convergence and diversity property. To achieve convergence, most multi-objective evolutionary algorithms today employ Pareto-ranking as the main criteria for fitness calculation. The computation of Pareto-rank in a population is time consuming, and arguably the most computationally expensive component in an iteration of the said algorithms. This paper proposes a Multi-objective Evolutionary Algorithm which avoids Pareto-ranking altogether by employing the transitivity of the domination relation. The proposed algorithm is an elitist algorithm with explicit diversity preservation procedure. It applies a measure reflecting the degree of domination between solutions in a steady-state replacement strategy to determine which individuals survive to the next iteration. Results on nine standard test functions demonstrated that the algorithm performs favorably compared to the popular NSGA-II in terms of convergence as well as diversity of the Pareto-set approximation, and is computationally more efficient.

This paper presents an evolutionary algorithm to develop cooperative strategies for power buyers in a deregulated electrical power market. Cooperative strategies are evolved through the collaboration of the buyer with other buyers defined by the different group memberships. The paper explores how buyers can lower their costs by using the algorithm that evolves their group sizes and memberships. The algorithm interfaces with PowerWorld Simulator to include in the technical aspect of a power system network, particularly the effects of the network constraints on the power flow. Simulation tests on an IEEE 14-bus transmission network are conducted and power buyer strategies are observed and analyzed.

This paper proposes a novel technique for automatic incident detection on highways using a hybrid neuro-fuzzy system. The proposed neuro-fuzzy system employs a self rule generating algorithm that organizes the training data into clusters and automatically learns the fuzzy rules. Modified linear least squares regression models are employed for training of parameters. Real I-880 freeway traffic data is used to test the effectiveness of the proposed algorithm. The results obtained show high potential for the application of this neurofuzzy system to automated traffic incident detection.

Real-time traffic signal control is an integral part of the urban traffic control system, and providing effective real-time traffic signal control for a large complex traffic network is an extremely challenging distributed control problem. This paper adopts the multiagent system approach to develop distributed unsupervised traffic responsive signal control models, where each agent in the system is a local traffic signal controller for one intersection in the traffic network. The first multiagent system is developed using hybrid computational intelligent techniques. Each agent employs a multistage online learning process to update and adapt its knowledge base and decision-making mechanism. The second multiagent system is developed by integrating the simultaneous perturbation stochastic approximation theorem in fuzzy neural networks (NN). The problem of real-time traffic signal control is especially challenging if the agents are used for an infinite horizon problem, where online learning has to take place continuously once the agent-based traffic signal controllers are implemented into the traffic network. A comprehensive simulation model of a section of the Central Business District of Singapore has been developed using PARAMICS microscopic simulation program. Simulation results show that the hybrid multiagent system provides significant improvement in traffic conditions when evaluated against an existing traffic signal control algorithm as well as the SPSA-NN-based multiagent system as the complexity of the simulation scenario increases. Using the hybrid NN-based multiagent system, the mean delay of each vehicle was reduced by 78% and the mean stoppage time, by 85% compared to the existing traffic signal control algorithm. The promising results demonstrate the efficacy of the hybrid NN-based multiagent system in solving large-scale traffic signal control problems in a distributed manner

This paper proposes a new hybrid neural network (NN) model that employs a multistage online learning process to solve the distributed control problem with an infinite horizon. Various techniques such as reinforcement learning and evolutionary algorithm are used to design the multistage online learning process. For this paper, the infinite horizon distributed control problem is implemented in the form of real-time distributed traffic signal control for intersections in a large-scale traffic network. The hybrid neural network model is used to design each of the local traffic signal controllers at the respective intersections. As the state of the traffic network changes due to random fluctuation of traffic volumes, the NN-based local controllers will need to adapt to the changing dynamics in order to provide effective traffic signal control and to prevent the traffic network from becoming overcongested. Such a problem is especially challenging if the local controllers are used for an infinite horizon problem where online learning has to take place continuously once the controllers are implemented into the traffic network. A comprehensive simulation model of a section of the Central Business District (CBD) of Singapore has been developed using PARAMICS microscopic simulation program. As the complexity of the simulation increases, results show that the hybrid NN model provides significant improvement in traffic conditions when evaluated against an existing traffic signal control algorithm as well as a new, continuously updated simultaneous perturbation stochastic approximation-based neural network (SPSA-NN). Using the hybrid NN model, the total mean delay of each vehicle has been reduced by 78% and the total mean stoppage time of each vehicle has been reduced by 84% compared to the existing traffic signal control algorithm. This shows the efficacy of the hybrid NN model in solving large-scale traffic signal control problem in a distributed manner. Also, it indicates the possibility of using the hybrid NN model for other applications that are similar in nature as the infinite horizon distributed control problem.

This paper presents an evolutionary algorithm, guided by heuristics, to solve the unit commitment and dispatch problem in large scale power systems. Unit commitment is a non linear, large scale, and with a varied set of constraints optimization problem for which there exist no exact solution techniques with a reasonable computation time. Problem-specific heuristics have been included to increase the speed of convergence and the efficiency of the algorithm. The initial random population was seeded with good solutions using a priority list method, and a problem specific genetic operator was used. A comparison of results with other solution techniques shows superior results, even on large scale systems.

In this paper a hybrid fuzzy evolutionary algorithm for a multi-objective resource allocation problem, the student project allocation (SPA) problem, is presented. Student project allocation must satisfy a number of soft objectives stemming from multiple points of view. The proposed algorithm employs a fuzzy inference system to model and aggregate the objectives, assuming the role of the fitness function in the evolutionary algorithm. The fuzzy system captures preferences of the decision maker in the compromise between various objectives, thereby guiding the search to interesting regions in the objective space. The results demonstrate the effectiveness of this hybrid approach for a large data set.

There are many learning algorithms available in the field of pattern classification and people are still discovering new algorithms that they hope will work better. Any new learning algorithm, beside its theoretical foundation, needs to be justified in many aspects including accuracy and efficiency when applied to real life problems. In this paper, we report the empirical comparison of a recent algorithm RM, its new extensions and three classical classifiers in different aspects including classification accuracy, computational time and storage requirement. The comparison is performed in a standardized way and we believe that this would give a good insight into the algorithm RM and its extension. The experiments also show that nominal attributes do have an impact on the performance of those compared learning algorithms.

The complexity involved in implementing multi-objective routing in computer networks has led to many researchers exploring alternate solutions with the use of heuristic based techniques. The rationale underlying the use of heuristic based priorities in achieving multiple objectives appears to be ad hoc and unclear due to the complex interactions among the various objectives. However these uncertainties can be effectively modeled using fuzzy set theory. This work introduces the notion of multi-objective route selection in mobile ad hoc networks (MANET) using a evolutionary fuzzy cost function to deliberately calculate cost adaptively. The fuzzy cost function is a continuous function of the metrics describing the state of a route. Simulation results demonstrate the superiority of the proposed technique over conventional MANET routing schemes.

Power converters are non-linear systems that usually employ linear controllers designed to offer good small signal performance at the nominal operating point. Yet, the converter's large-signal response is generally poor. Fuzzy logic controllers (FLCs) used in such cases improve the response, but their small-signal response is not as good as that of linear controller. In this paper, design of a SISO-FLC that offers good small and large-signal performance in a boost converter is presented. Besides reducing control complexity, the FLC offers better transient response in the converter than the linear controller. Simulation results are presented to show this. Using describing function method, the system stability margins are also analyzed.

We summarize some of the key research areas in brain sciences and neural networks that have recently been or are being worked on by researchers in Singapore. Researchers in Singapore are developing theory of neural networks, notably improved radial basis function networks, fuzzy neural networks, and fast learning neural networks. Applications of neural networks include bioinformatics, multimedia, data mining, and communications. Researchers are also working with neurophysiologists on functional brain imaging and brain disease analysis.

A novel method using a reduced multivariate polynomial model has been developed for biometric decision fusion where simplicity and ease of use could be a concern. However, much to our surprise, the reduced model was found to have good classification accuracy for several commonly used data sets from the Web. In this paper, we extend the single output model to a multiple outputs model to handle multiple class problems. The method is particularly suitable for problems with small number of features and large number of examples. Basic component of this polynomial model boils down to construction of new pattern features which are sums of the original features and combination of these new and original features using power and product terms. A linear regularized least-squares predictor is then built using these constructed features. The number of constructed feature terms varies linearly with the order of the polynomial, instead of having a power law in the case of full multivariate polynomials. The method is simple as it amounts to only a few lines of Matlab code. We perform extensive experiments on this reduced model using 42 data sets. Our results compared remarkably well with best reported results of several commonly used algorithms from the literature. Both the classification accuracy and efficiency aspects are reported for this reduced model.

Automated incident detection is an essential component of a modern freeway traffic monitoring system. A number of neural network (NN)-based incident detection models have been tested independently over the past decade. This paper evaluates the adaptability of three promising NN models for this problem: a multilayer feed-forward NN (MLFNN), a basic probabilistic NN (BPNN) and a constructive probabilistic NN (CPNN). These three models have been developed on an original freeway site in Singapore and then adapted to a new freeway site in California. In addition to their incident detection performance, their ability to adapt to new freeway sites, and network sizes have also been compared. A novel updating scheme has been used for adjustment of smoothing parameter of the BPNN. Results of this study show that the MLFNN model has the best incident detection performance at the development site while CPNN model has the best performance after model adaptation at the new site. In addition, the adaptation method for CPNN model is less laborious. The efficient network pruning procedure for the CPNN network resulted in a smaller network size, making it easier to implement it for real-time application. The results suggest that CPNN model has good potential for application in an operational automatic incident detection system for freeways.

We describe particle swarm inspired evolutionary algorithm (PS-EA), which is a hybridized evolutionary algorithm (EA) combining the concepts of EA and particle swarm theory. PS-EA is developed in aim to extend PSO algorithm to effectively search in multiconstrained solution spaces, due to the constraints rigidly imposed by the PSO equations. To overcome the constraints, PS-EA replaces the PSO equations completely with a self-updating mechanism (SUM), which emulates the workings of the equations. A comparison is performed between PS-EA with genetic algorithm (GA) and PSO and it is found that PS-EA provides an advantage over typical GA and PSO for complex multimodal functions like Rosenbrock, Schwefel and Rastrigrin functions. An application of PS-EA to minimize the classic Fonseca 2-objective functions is also described to illustrate the feasibility of PS-EA as a multiobjective search algorithm.

This paper introduces a particle swarm optimization-based method for solving a multi-objective generator maintenance scheduling problem with many constraints. It is shown that the particle swarm optimization-based approach is effective in obtaining feasible schedules in a reasonable time. Actual data from a practical power system was used in this study and results were compared against those from other evolutionary methods on the same set of data. This paper also introduces a novel concept for the spawning and selection mechanism in a hybrid particle swarm algorithm. The results suggest that this hybrid model converges to a better solution faster than the standard PSO algorithm. It is envisaged that this hybrid approach can be easily implemented for similar optimization and scheduling problems to obtain better convergence.

This paper proposes a new approach to automatic incident detection on traffic highways using particle swarm optimization (PSO). The rampant growth in traffic incidents, which is high cost incurring, has led to significant interest in the development of effective incident detection techniques in recent years. Various techniques have been proposed to effectively address this problem, the most promising of which are artificial neural networks (ANN) based methods. Backpropagation (BP) has proven to be one of the best methods to train weights of ANN for incident detection. However it has several limitations including slow convergence, heuristic determination of parameters and possibility of getting stuck in a local minima. This paper overcomes these problems by using particle swarm optimization to train a neural network in place of BP. Actual data from a highway was used for training and testing of this method. Simulation results show that PSO performed better than the backpropagation algorithm.

This paper presents a new hybrid, synergistic approach in applying computational intelligence concepts to implement a cooperative, hierarchical, multiagent system for real-time traffic signal control of a complex traffic network. The large-scale traffic signal control problem is divided into various subproblems, and each subproblem is handled by an intelligent agent with a fuzzy neural decision-making module. The decisions made by lower-level agents are mediated by their respective higher-level agents. Through adopting a cooperative distributed problem solving approach, coordinated control by the agents is achieved. In order for the multiagent architecture to adapt itself continuously to the dynamically changing problem domain, a multistage online learning process for each agent is implemented involving reinforcement learning, learning rate and weight adjustment as well as dynamic update of fuzzy relations using an evolutionary algorithm. The test bed used for this research is a section of the Central Business District of Singapore. The performance of the proposed multiagent architecture is evaluated against the set of signal plans used by the current real-time adaptive traffic control system. The multiagent architecture produces significant improvements in the conditions of the traffic network, reducing the total mean delay by 40% and total vehicle stoppage time by 50%.

Finding a feasible lecture/tutorial timetable in a large university department is a challenging problem faced continually in educational establishments. This paper presents an evolutionary algorithm (EA) based approach to solving a heavily constrained university timetabling problem. The approach uses a problem-specific chromosome representation. Heuristics and context-based reasoning have been used for obtaining feasible timetables in a reasonable computing time. An intelligent adaptive mutation scheme has been employed for speeding up the convergence. The comprehensive course timetabling system presented in this paper has been validated, tested and discussed using real world data from a large university.

This paper presents the application of fuzzy-neuro-evolutionary hybrid system with online reinforcement learning for intelligent road traffic management and control. Taking a step away from the conventional traffic control system, the hybrid system presents different methodologies in knowledge acquisition, decisionmaking, learning and goal formulation with the use of a three-layered hierarchical, distributed agent architecture. Distributed and hierarchical fuzzy knowledge acquisition allows different levels of perception to be derived for the same traffic situation by the intelligent agents. Agents' perceptions can be changed with the use of online reinforcement learning. Initial experimental results show that the implementation of the hybrid agents in the traffic network generally yields better network performance when compared to a network without the agents. The probability of a traffic network evolving into pathological states with oversaturation is also reduced with the implementation of the agents.

Presents the development of a universal fuzzy controller for a non-linear power electronic converter. The classical boost converter used in power supplies is a minimal phase system with a right-half-plane zero. This right-half-plane zero forces the designer to go for controllers that give slow dynamics. The conventional linear PI controllers for such converters, designed under the worst case conditions of maximum load and minimum line conditions, present a lower loop bandwidth even if the operating conditions are better. Moreover, the system response is sluggish. Modern fuzzy controllers, on the other hand, can be designed to adapt to varying operating conditions for application in such nonlinear systems. The paper designs a universal fuzzy controller and compares its performance at various operating points with local PI controllers designed for the particular operating points. The settling time and overshoot for startup and step response obtained by computer simulations have been compared. The superior performance of the fuzzy controller, in particular, its ability to achieve good transient response under different operating conditions is clearly established.

A novel routing scheme for mobile ad hoc networks (MANETs), which combines the on-demand routing capability of Ad Hoc On-Demand Distance Vector (AODV) routing protocol with a distributed topology discovery mechanism using ant-like mobile agents is proposed in this paper. The proposed hybrid protocol reduces route discovery latency and the end-to-end delay by providing high connectivity without requiring much of the scarce network capacity. On the one side the proactive routing protocols in MANETs like Destination Sequenced Distance Vector (DSDV) require to know, the topology of the entire network. Hence they are not suitable for highly dynamic networks such as MANETs, since the topology update information needs to be propagated frequently throughout the network. These frequent broadcasts limit the available network capacity for actual data communication. On the other hand, on-demand, reactive routing schemes like AODV and Dynamic Source Routing (DSR), require the actual transmission of the data to be delayed until the route is discovered. Due to this long delay a pure reactive routing protocol may not be applicable for real-time data and multimedia communication. Through extensive simulations in this paper it is proved that the proposed Ant-AODV hybrid routing technique, is able to achieve reduced end-to-end delay compared to conventional ant-based and AODV routing protocols.

This paper proposes a new technique for freeway incident detection using a constructive probabilistic neural network (CPNN). The CPNN incorporates a clustering technique with an automated training process. The work reported in this paper was conducted on Ayer Rajah Expressway (AYE) in Singapore for incident detection model development, and subsequently on I-880 freeway in California, for model adaptation. The model developed achieved incident detection performance of 92% detection rate and 0.81% false alarm rate on AYE, and 91.30% detection rate and 0.27% false alarm rate on I-880 freeway using the proposed adaptation method. In addition to its superior performance, the network pruning method employed facilitated model size reduction by a factor of 11 compared to a conventional probabilistic neural network. A more impressive size reduction by a factor of 50 was achieved after the model was adapted for the new site. The results from this paper suggest that CPNN is a better adaptive classifier for incident detection problem with a changing site traffic environment.