Recent advances in data-driven methods have significantly improved battery capacity estimation, yet most existing approaches remain constrained by their reliance on supervised learning, requiring substantial amounts of labeled cycling data that are often costly to obtain. To address this challenge, this study proposes a dual-branch network-based semi supervised framework that integrates self-supervised learning and transfer learning mechanisms. First, the framework derives meaningful degradation-aware auxiliary labels from both labeled and unlabeled samples, creating reliable self-supervised signal for model training. Second, the designed dual-branch neural network architecture employs a shared feature extractor that processes input data for both the primary capacity estimation task and the auxiliary label prediction task, enabling effective knowledge transfer between labeled and unlabeled domains through their common representation space. Third, a pseudolabel filtering strategy is proposed to dynamically select high-confidence samples from the unlabeled dataset for self-training, thereby effectively expanding the training set with high-quality pseudolabels and enhancing the capacity estimation accuracy. Finally, extensive experiments validate the framework’s superior performance, achieving a worst-case root-mean-square error of only 0.0143 Ah with merely 5% labeled data, representing a 27.04% reduction compared with the best-performing semi supervised baseline (Co-training) under the same conditions.

High reliability and low latency are essential to railway wireless communications, which transmit train control and dispatch commands to ensure operational safety. However, as railway systems become increasingly electrified and more complex, the exposure to electromagnetic interference (EMI) also grows, potentially causing service disruptions and compromising safety. Intentional EMI (IEMI), which is deliberately and often maliciously generated, further increases the vulnerability of these critical communication networks. Real-time detection and classification of EMI and IEMI therefore become increasingly important. This paper presents composite models that reflect realistic railway scenarios and proposes an adaptive classification approach for EMI and IEMI using a deep learning algorithm based on bidirectional long-short-term memory (BiLSTM) networks and attention mechanisms. By employing time-series feature extraction to analyze both time and frequency information at fine resolution, the proposed method demonstrates a classification accuracy of 94.98%. Simulation results outperform existing techniques with a 3% improvement in accuracy, showcasing its adaptability across four typical railway scenarios at train speeds of up to 500 km/h. Moreover, online monitoring phase performs real-time detection in just 7.43 ms, meeting the stringent latency requirements for railway systems. Validation using real-world data further confirms the practical applicability of the proposed methods under actual operating conditions.

Plug-and-play (PnP) operations of distributed generation units (DGUs) with constant power loads (CPLs) often destabilize dc microgrids (DCmGs). To address this issue, this article proposes a scalable neural network control strategy for nonlinear DCmGs with CPLs, enabling seamless PnP operations of DGUs. A radial basis function neural network is employed to handle the uncertain CPL nonlinearity without requiring any prior knowledge. A structured Lyapunov matrix is utilized to eliminate the coupling effects of power lines by reshaping them into a Laplacian matrix structure. Within this framework, a scalable neural network control approach is proposed, integrating a nominal controller with explicit gain inequalities and an adaptive controller governed by an adaptation law. This approach operates locally, independent of other DGUs and power lines, ensuring PnP operations and maintaining uniformly ultimately bounded stability. The effectiveness of the proposed method is validated through case studies on a modified IEEE 37-bus test system.

The proliferation of high-speed rail (HSR) networks and railway electrification has advanced the integration of the latest wireless communication networks with railway systems. Ensuring a reliable bidirectional communication link between moving trains and base stations is crucial to maintaining the safety of real-time rail operations. However, the growing complex-ity of railway systems and increased exposure to electromagnetic emissions present substantial challenges. In particular, railway wireless communication networks are vulnerable to various kinds of electromagnetic interference (EMI) and intentional EMI (IEMI), which could cause operational disruptions and safety hazards. This paper proposes a real-time classification method for EMI and IEMI, using deep learning-based bidirectional long-short-term memory (BiLSTM) networks. By employing multivariate time-series characteristics, the method can simul-taneously learn both time and frequency information at a finer resolution, offering better performance than existing methods. The simulation results demonstrate a high accuracy of 93.4% and adaptability at different speeds and in various scenarios.

The plugging-in/-out of renewable distributed generation units (DGUs) often alters the microgrid size and coupling terms, resulting in computational burdens and voltage shocks. This article proposes a novel scalable fuzzy voltage control scheme for nonlinear direct current microgrids (DCmGs) composed of DGUs and constant power loads (CPLs) interconnected via power lines. First, a Takagi–Sugeno fuzzy DCmG model with CPL is formulated to capture nonlinear characteristics and diverse transient behaviors. Then, a scalable fuzzy control approach is developed to mitigate negative coupling effects of power lines. This is achieved by a novel argument that leverages dissipativity theory to transform such effects into linear matrix inequalities and subsequently imposes constraints on their sequential principal minors. Specifically, the proposed control method operates locally and independently of other DGUs and line couplings, enabling seamless plug-and-play operations without updating any controllers. Finally, theoretical results are validated through simulations using the MATLAB/SimPowerSystems toolbox.

Indoor temperature prediction of multiple zones in near future horizons is vital in developing an optimal regulation strategy of heating, ventilation, and air conditioning systems in large-scale complex buildings. This is, however, challenging due to the spatial–temporal correlation and multivariable coupling characteristics. This article proposes a novel deep learning framework incorporating the distributed long short-term memory and graph convolution network namely DL-GCN for indoor temperature prediction in large public buildings, aiming to learn the spatial–temporal correlation and multivariable coupling features. First, the indoor temperature and humidity data from different zones are handled by GCN networks to extract the temperature spatial features. Then, in the distributed LSTM module, other data, such as light and ac power consumption, are fused with the outputs of the GCN module, respectively, in a distributed way to learn the coupling interactions and temporal characteristics between these variables. Comparison study and ablation experiments are conducted using real datasets from a large-scale building to verify its effectiveness and superior performance in multizone indoor temperature prediction.

A deep transfer learning method is presented for establishing the aggregated system frequency response (SFR) model of wind-thermal hybrid power systems (HPSs). In order to deal with nonlinearities and non-Gaussian disturbances, the quadratic survival information potential of the squared identification error is employed to construct the performance index when training recurrent neural networks (RNNs). A pretrained SFR model is then obtained by the improved RNNs using the source domain data collected from the HPS in historical scenarios. Subsequently, the maximum mean difference is utilized to test the similarity of the HPS in historical and current scenarios. After that, the pretrained SFR model is fine-tuned by adding some nodes to the recurrent layer and a functional link to the input layer. The SFR model of the HPS operating in current scenario can, then, be built based on the transferred source domain pretrained SFR model. Simulation results illustrate that the proposed data driven modeling method can obtain accurate, effective and timely SFR model for a wind-thermal HPS with different wind speeds and load disturbances.

Multioutput regression of nonlinear and nonstationary data is largely understudied in both machine learning and control communities. This article develops an adaptive multioutput gradient radial basis function (MGRBF) tracker for online modeling of multioutput nonlinear and nonstationary processes. Specifically, a compact MGRBF network is first constructed with a new two-step training procedure to produce excellent predictive capacity. To improve its tracking ability in fast time-varying scenarios, an adaptive MGRBF (AMGRBF) tracker is proposed, which updates the MGRBF network structure online by replacing the worst performing node with a new node that automatically encodes the newly emerging system state and acts as a perfect local multioutput predictor for the current system state. Extensive experimental results confirm that the proposed AMGRBF tracker significantly outperforms existing state-of-the-art online multioutput regression methods as well as deep-learning-based models, in terms of adaptive modeling accuracy and online computational complexity.

With consideration of false data injection (FDI) on the demand side, it brings a great challenge for the optimal defensive strategy with the security issue, voltage stability, power flow, and economic cost indexes. This article proposes a Takagi-Sugeuo-Kang (TSK) fuzzy system-based reinforcement learning approach for the resilient optimal defensive strategy of interconnected microgrids. Due to FDI uncertainty of the system load, TSK-based deep deterministic policy gradient (DDPG) is proposed to learn the actor network and the critic network, where multiple indexes' assessment occurs in the critic network, and the security switching control strategy is made in the actor network. Alternating direction method of multipliers (ADMM) method is improved for policy gradient with online coordination between the actor network and the critic network learning, and its convergence and optimality are proved properly. On the basis of security switching control strategy, the penalty-based boundary intersection (PBI)-based multiobjective optimization method is utilized to solve economic cost and emission issues simultaneously with considering voltage stability and rate-of-change of frequency (RoCoF) limits. According to simulation results, it reveals that the proposed resilient optimal defensive strategy can be a viable and promising alternative for tackling uncertain attack problems on interconnected microgrids.

Due to uncertainty and dynamic characteristics from intermittent energy and load demand response (DR), the optimal operation of the hybrid energy system is a great challenge. This article proposes an event-triggered multiagent coordinated optimization strategy with two-layered architecture. First, the price-bidding-based DR model is proposed with different stakeholders, and it also deduces the optimal bidding price with the Nash equilibrium theory. Then, four agents are designed to control different kinds of energy resources: agent 1 mainly analyzes the uncertainty or randomness caused by intermittent power, agent 2 takes charge of the dynamic economic dispatch (DED) within thermal units, agent 3 manages the optimal scheduling of energy storage, and agent 4 mainly undertakes the load-shifting strategy from consumers. In the upper-layer level, all agents coordinate together to ensure the stability of the hybrid energy system with an event-triggered mechanism, and the intelligent control approach mainly depends on switching ON/OFF power generators or curtailing system load, and the consensus algorithm is utilized to optimize the subsystem problem in the lower-layer level. Furthermore, the simulation results can further verify the efficiency of the proposed method, and it also reveals that the event-triggered multiagent optimization strategy can be a promising way to solve the hybrid energy system problem.

The stochastic nature of intermittent energy resources has brought significant challenges to the optimal operation of the hybrid energy systems. This article proposes a probabilistic multiobjective evolutionary algorithm based on decomposition (MOEA/D) method with two-step risk-based decision-making strategy to tackle this problem. A scenario-based technique is first utilized to generate a stochastic model of the hybrid energy system. Those scenarios divide the feasible domain into several regions. Then, based on the MOEA/D framework, a probabilistic penalty-based boundary intersection (PBI) with gradient descent differential evolution (GDDE) algorithm is proposed to search the optimal scheme from these regions under different uncertainty budgets. To ensure reliable and low risk operation of the hybrid energy system, the Markov inequality is employed to deduce a proper interval of the uncertainty budget. Further, a fuzzy grid technique is proposed to choose the best scheme for real-world applications. The experimental results confirm that the probabilistic adjustable parameters can properly control the uncertainty budget and lower the risk probability. Further, it is also shown that the proposed MOEA/D-GDDE can significantly enhance the optimization efficiency.

Industry 4.0 applications rely upon timely and accurate data about plant and process within a production site. Whilst modern facilities tend to have this capability as a matter of course, older equipment may lack network connectivity. A lack of data-gathering capability represents a significant barrier-to-entry when undertaking any data-driven investigation or improvement programs. Wireless sensor networks (WSNs) can be used as a flexible and low-disruption technique to acquire data at the point of interest, however the data stream is often lossy when deployed in harsh conditions without costly adaptations to the environment.This paper introduces the F-QoS metric which is able to classify the quality of the data stream from a WSN (using only packet reception timestamps), at user-defined sampling rates with a constraint placed upon the maximum amount of missing data. The resulting classifications can be used in an offline fashion to select periods of high-quality data for modelling, or, in an online manner to assess the realtime performance of a WSN.The F-QoS metric is applied to a LoRaWAN network in a large commercial bakery with a low-disruption installation-the network links are strained by large metal obstructions and the endpoints are installed inside metal cabinets. Each node transmits on a 10s cycle, and the analysis shows that >70% of the data is suitable for sampling at a 30s rate. The results indicate that LoRaWAN is capable of data acquisition in an unadapted and challenging environment, with the recommendation that the raw sample rate should be triple the desired final sample rate.

Fault feature extraction based on prior knowledge and raw data is increasingly becoming more challenging in energy Internet fault diagnosis due to complicated network topology and coupling disturbances introduced into the systems. Deep learning methods that have emerged in recent years, such as the convolutional neural networks (CNNs), have shown a number of advantages and great potentials in the field of feature extraction and image recognition. However, CNNs does not work well in fault diagnosis for industrial systems, due to the totally different data representations between images used in recognition and signals obtained from industrial processes. This paper tackles this problem by introducing a novel and generic fault diagnosis method for complicated system, namely, the Spearman rank correlation-based CNNs (SR-CNNs). By imposing the Spearman rank correlation image layer on the typical CNNs, the multiple time-series signals measured by the phasor measurement units (PMUs) is converted to appropriate data images, which are then fed to the CNNs. With the aid of this novel design, different fault features can be comprehensively extracted while the fault can be identified more quickly and precisely than other conventional approaches. To validate the efficacy of the proposed approach, an IEEE defined power gird with many new energy resources are used as the test platform. The experimental results confirm the effectiveness and superiority of the proposed method in energy Internet fault diagnosis over conventional methods.

As a novel family member of the redox flow batteries (RFBs), the single flow zinc-nickel battery (ZNB) without ion exchange membranes has attracted a lot of interests in recent years due to the high charging and discharging efficiencies. To understand the electrical behaviour is a key for proper battery management system. Unlike the electrochemical mechanism models and equivalent circuit models, the neural network based black-box model does not need knowledge about the electrochemical reactions and is a promising and adaptive approach for the ZNB battery modelling. In this paper, a compact radial basis function neural network is developed using a two-stage layer selection strategy to determine the network structure. While Jaya optimization is utilized to determine the non-linear parameters in the selected hidden nodes of the resultant RBF neural network (RBF-NN) model. The proposed method is implemented to model the ZNB to capture the non-linear electric behaviours through the readily measurable input signals. Experimental results manifest the accurate prediction capability of the resultant neural model and confirm the effectiveness of the proposed approach.

This paper applies advanced battery modeling and multiobjective constrained nonlinear optimization techniques to derive suitable charging patterns for lithium-ion batteries. Three important yet competing charging objectives, including battery health, charging time, and energy conversion efficiency, are taken into account simultaneously. These optimization objectives are first subject to a high-fidelity battery model that is synthesized from recently developed individual electrical, thermal, and aging models. The coupling relationship and multiple timescales among different model dynamics are identified. Furthermore, constraints are imposed explicitly on the current, voltage, state-of-charge, and temperature. Such a complex charging problem is solved by using an ensemble multiobjective biogeography-based optimization approach. As a result, two charging patterns, namely the constant current-constant voltage (CC-CV) and multistage CC-CV, are optimized to balance various combinations of charging objectives. Different tradeoffs and sensitive elements are compared and analyzed based on the Pareto frontiers. Illustrative results demonstrate that the proposed strategy can effectively offer feasible health-conscious charging with desirable tradeoffs among charging speed and energy conversion efficiency under different demand priorities.

Ship trajectory length prediction is vital for intelligent traffic signaling in the controlled waterways of the Yangtze River. In current intelligent traffic signaling systems (ITSSs), ships are supposed to travel exactly along the central line of the Yangtze River, which is often not a valid assumption and has caused a number of problems. Over the past few years, traffic data have been accumulated exponentially, leading to the big data era. This trend allows more accurate prediction of ships' travel trajectory length based on historical data. In this paper, ships' historical trajectories are first grouped by using the fuzzy c-means clustering algorithm. The relationship between some known factors (i.e., ship speed, loading capacity, self-weight, maximum power, ship length, ship width, ship type, and water level) and the resultant memberships are then modeled using artificial neural networks. The trajectory length is then estimated by the sum of the predicted probabilities multiplied by the trajectory cluster centers' length. To the best of our knowledge, this is the first time to predict the overall trajectory length of manually controlled ships. The experimental results show that the proposed method can reduce the probability of generating incorrect traffic control signals by 74.68% over existing ITSSs. This will significantly improve the efficiency of the Yangtze River traffic management system and increase the traffic capacity by reducing the traveling time.

Yangtze River is probably the world's busiest inland waterway. Ships need to be guided when passing through a controlled waterway based on their long-term speed prediction. Inaccurate ship speed prediction leads to nonoptimal traffic signaling, which may cause a significant traffic jam. For the existing intelligent traffic signaling system, the ship speed is assumed to be constant, which has caused many problems and issues. This paper proposes a novel algorithm to construct an improved multilayer perceptron (MLP) network for accurate long-term ship speed prediction, in which the hidden neurons of the MLP are optimized by the particle swarm optimization method. The effectiveness and efficiency of the method are guaranteed by using the orthogonal least squares method, which is the fast approach for the construction of the MLP network in a stepwise forward procedure. The model is driven by easily acquired dynamic data of the ships, including the speed and the position. The effectiveness of the proposed method is further confirmed by comparing with several traditional modeling techniques. To the best of our knowledge, this is the first time that a ship speed model is built for long-term prediction. The experimental results show that the developed model is in good agreement with the real-life data, with more than 97% accuracy. It will help to generate the optimal traffic commands for Yangtze River in an intelligent traffic signaling system.

Battery charging strategy is a key issue in battery management system to ensure good battery performance and safe operation during the charging process. In this paper, a novel battery optimal charging strategy is proposed by applying the TLBO algorithm to a LiFeP04 battery for an optimal charging based on a coupled thermoelectric model. A specific dual-objective function including battery charging time and temperature rise (both battery interior and surface) is formulated first. Then a battery optimal charging strategy is presented in detail by using the TLBO algorithm, aiming at finding a suitable constant-current-constant-voltage (CCCV) current profile to minimize the dual-objective function. Besides, the effects of different weights in dual-objective function on the optimal charging profile are analyzed. Simulation results demonstrate that the presented optimal charging strategy can provide effective and acceptable optimal charge current profile. The strategy can be also easily implemented to other battery types to effectively balance the battery charging time and battery temperature rise during charging process.

Unit commitment is a key issue in power system operation and has long been an intractable problem due to its complex mix-integer nonlinear formulation. The original unit commitment problem aims to minimize the fossil fuel cost by determining the on/off status of power units and power contribution of each online unit at the same time. However, the uncoordinated large charging power necessity of plug-in electric vehicles brings unprecedented challenges to the power system operators and further complicates the unit commitment problem. To seamless integrate the plug-in electric vehicles into the unit commitment, a new binary/real-value hybrid meta-heuristic algorithm framework is proposed in this paper, simultaneously determining the binary status and power output of units as well as the power delivered to/feedback from flexible charging and discharging of plug-in electric vehicles. A batch of binary particle swarm optimisation variants with different transfer functions are implemented and compared in solving the unit commitment problem with and without plug-in electric vehicles. Numerical studies illustrate the effectiveness of the proposed intelligent algorithm and the impact of different transfer functions is evaluated.

We present a new wrapper feature selection algorithm for human detection. This algorithm is a hybrid featureselection approach combining the benefits of filter and wrapper methods. It allows the selection of an optimalfeature vector that well represents the shapes of the subjects in the images. In detail, the proposed featureselection algorithm adopts the k-fold subsampling and sequential backward elimination approach, while thestandard linear support vector machine (SVM) is used as the classifier for human detection. We apply theproposed algorithm to the publicly accessible INRIA and ETH pedestrian full image datasets with the PASCALVOC evaluation criteria. Compared to other state of the arts algorithms, our feature selection based approachcan improve the detection speed of the SVM classifier by over 50% with up to 2% better detection accuracy.Our algorithm also outperforms the equivalent systems introduced in the deformable part model approach witharound 9% improvement in the detection accuracy

Electric vehicles provide an opportunity to reduce fossil fuel consumptions and to decrease the emissions of green-house gas and air pollutants from the transport sector. The adoption of a large number of plug-in electric vehicles however imposes significant impacts on the power system operation due to uncertain charging and discharging patterns. In this paper, multiple charging and discharging scenarios of electric vehicles together with the grid integration of renewable energy sources are examined and evaluated within the unit commitment problem. A quantum-inspired binary particle swarm optimization method is employed to determine the on/off status of each unit. Comparative studies show that the off-peak charging and peak discharging scenario is a viable option to significantly reduce the economic cost and to complement the renewable energy generation.

A number of neural networks can be formulated as the linear-in-the-parameters models. Training such networks can be transformed to a model selection problem where a compact model is selected from all the candidates using subset selection algorithms. Forward selection methods are popular fast subset selection approaches. However, they may only produce suboptimal models and can be trapped into a local minimum. More recently, a two-stage fast recursive algorithm (TSFRA) combining forward selection and backward model refinement has been proposed to improve the compactness and generalization performance of the model. This paper proposes unified two-stage orthogonal least squares methods instead of the fast recursive-based methods. In contrast to the TSFRA, this paper derives a new simplified relationship between the forward and the backward stages to avoid repetitive computations using the inherent orthogonal properties of the least squares methods. Furthermore, a new term exchanging scheme for backward model refinement is introduced to reduce computational demand. Finally, given the error reduction ratio criterion, effective and efficient forward and backward subset selection procedures are proposed. Extensive examples are presented to demonstrate the improved model compactness constructed by the proposed technique in comparison with some popular methods.

One of the main purposes of building a battery model is for monitoring and control during battery charging/discharging as well as for estimating key factors of batteries such as the state of charge for electric vehicles. However, the model based on the electrochemical reactions within the batteries is highly complex and difficult to compute using conventional approaches. Radial basis function (RBF) neural networks have been widely used to model complex systems for estimation and control purpose, while the optimization of both the linear and non-linear parameters in the RBF model remains a key issue. A recently proposed meta-heuristic algorithm named Teaching-Learning-Based Optimization (TLBO) is free of presetting algorithm parameters and performs well in non-linear optimization. In this paper, a novel self-learning TLBO based RBF model is proposed for modelling electric vehicle batteries using RBF neural networks. The modelling approach has been applied to two battery testing data sets and compared with some other RBF based battery models, the training and validation results confirm the efficacy of the proposed method.

To tackle the twin challenges of sustainable energy supply and climate change, numerous efforts have been made to decarbonize the whole energy systems. Control Engineering, which concerns the automated operation of a machine or system to achieve desired target(s) and to avoid unstable or unintended disruptive behavior, has played a key role in modern industry and across the whole energy system. Advanced control technologies, such as optimal control, provide a framework to simultaneously regulate the system performance and limit control energy. However, little has been done so far to exploit the full potential of controller design in reducing the energy consumption while maintaining desirable system performance. This paper for the first time investigates the correlation between control energy consumption and system performance, and shows that this correlation is nonlinear and the controller design should be a delicate synthesis procedure to achieve better trade-off between system performance and energy saving.

The Takagi–Sugeno–Kang-type rule-based fuzzy model has found many applications in different fields; a major challenge is, however, to build a compact model with optimized model parameters which leads to satisfactory model performance. To produce a compact model, most existing approaches mainly focus on selecting an appropriate number of fuzzy rules. In contrast, this paper considers not only the selection of fuzzy rules but also the structure of each rule premise and consequent, leading to the development of a novel compact rule-based fuzzy model. Here, each fuzzy rule is associated with two sets of input attributes, in which the first is used for constructing the rule premise and the other is employed in the rule consequent. A new hybrid learning method combining the modified harmony search method with a fast recursive algorithm is hereby proposed to determine the structure and the parameters for the rule premises and consequents. This is a hard mixed-integer nonlinear optimization problem, and the proposed hybrid method solves the problem by employing an embedded framework, leading to a significantly reduced number of model parameters and a small number of fuzzy rules with each being as simple as possible. Results from three examples are presented to demonstrate the compactness (in terms of the number of model parameters and the number of rules) and the performance of the fuzzy models obtained by the proposed hybrid learning method, in comparison with other techniques from the literature.

The majority of reported learning methods for Takagi-Sugeno-Kang (TSK) fuzzy neural models to date mainly focus on improvement of their accuracy. However, one of the key design requirements in building an interpretable fuzzy model is that each obtained rule consequent must match well with the system local behavior when all the rules are aggregated to produce the overall system output. This is one of the distinctive characteristics from black-box models such as neural networks. Therefore, how to find a desirable set of fuzzy partitions and, hence, identify the corresponding consequent models which can be directly explained in terms of system behavior, presents a critical step in fuzzy neural modeling. In this paper, a new learning approach considering both nonlinear parameters in the rule premises and linear parameters in the rule consequents is proposed. Unlike the conventional two-stage optimization procedure widely practiced in the field where the two sets of parameters are optimized separately, the consequent parameters are transformed into a dependent set on the premise parameters, thereby enabling the introduction of a new integrated gradient descent learning approach. Thus, a new Jacobian matrix is proposed and efficiently computed to achieve a more accurate approximation of the cost function by using the second-order Levenberg-Marquardt optimization method. Several other interpretability issues regarding the fuzzy neural model are also discussed and integrated into this new learning approach. Numerical examples are presented to illustrate the resultant structure of the fuzzy neural models and the effectiveness of the proposed new algorithm, and compared with the results from some well-known methods.

The construction of a radial basis function (RBF) network involves the determination of the model size, hidden nodes, and output weights. Least squares-based subset selection methods can determine a RBF model size and its parameters simultaneously. Although these methods are robust, they may not achieve optimal results. Alternatively, gradient methods are widely used to optimize all the parameters. The drawback is that most algorithms may converge slowly as they treat hidden nodes and output weights separately and ignore their correlations. In this paper, a new discrete-continuous algorithm is proposed for the construction of a RBF model. First, the orthogonal least squares (OLS)-based forward stepwise selection constructs an initial model by selecting model terms one by one from a candidate term pool. Then a new Levenberg-Marquardt (LM)-based parameter optimization is proposed to further optimize the hidden nodes and output weights in the continuous space. To speed up the convergence, the proposed parameter optimization method considers the correlation between the hidden nodes and output weights, which is achieved by translating the output weights to dependent parameters using the OLS method. The correlation is also used by the previously proposed continuous forward algorithm (CFA). However, unlike the CFA, the new method optimizes all the parameters simultaneously. In addition, an equivalent recursive sum of squared error is derived to reduce the computation demanding for the first derivatives used in the LM method. Computational complexity is given to confirm the new method is much more computationally efficient than the CFA. Different numerical examples are presented to illustrate the effectiveness of the proposed method. Further, Friedman statistical tests on 13 classification problems are performed, and the results demonstrate that RBF networks built by the new method are very competitive in comparison with some popular classifiers.

This paper proposes a new methodology for solving the unmanned multi-vehicle formation control problem. It employs a unique “extension-decomposition-aggregation” scheme to transform the overall complex formation control problem to a group of sub-problems which work via boundary interactions. The H∞robust control strategy is applied to design the decentralised formation controllers to reject the interactions and work jointly to maintain the stability of the overall formation. Simulation studies have been performed to verify its performance and effectiveness.

Fuzzy-neural-network-based inference systems are well-known universal approximators which can produce linguistically interpretable results. Unfortunately, their dimensionality can be extremely high due to an excessive number of inputs and rules, which raises the need for overall structure optimization. In the literature, various input selection methods are available, but they are applied separately from rule selection, often without considering the fuzzy structure. This paper proposes an integrated framework to optimize the number of inputs and the number of rules simultaneously. First, a method is developed to select the most significant rules, along with a refinement stage to remove unnecessary correlations. An improved information criterion is then proposed to find an appropriate number of inputs and rules to include in the model, leading to a balanced tradeoff between interpretability and accuracy. Simulation results confirm the efficacy of the proposed method.

Recent years have witnessed an incredibly increasing interest in the topic of incremental learning. Unlike conventional machine learning situations, data flow targeted by incremental learning becomes available continuously over time. Accordingly, it is desirable to be able to abandon the traditional assumption of the availability of representative training data during the training period to develop decision boundaries. Under scenarios of continuous data flow, the challenge is how to transform the vast amount of stream raw data into information and knowledge representation, and accumulate experience over time to support future decision-making process. In this paper, we propose a general adaptive incremental learning framework named ADAIN that is capable of learning from continuous raw data, accumulating experience over time, and using such knowledge to improve future learning and prediction performance. Detailed system level architecture and design strategies are presented in this paper. Simulation results over several real-world data sets are used to validate the effectiveness of this method.

Learning from data streams has inspired considerable interests in recent years due to its wide applications in the fields such as network intrusion detection, credit fraud identification, spam filtering, and many others. Given the fact that most algorithms developed thus far assume the class distribution of the streaming data is relatively balanced, they will inevitably be confronted with severe performance deterioration when handling the imbalanced data streams. Evolved from our previous work SERA (SElectively Recursive Approach), the MuSeRA algorithm is proposed in this paper to deal with the problem of learning from imbalanced data streams. By maintaining an ensemble consisting of hypotheses built upon the coming training data chunks balanced by selectively accommodating previous minority examples, MuSeRA can efficiently learn the target concept of the imbalanced data streams and thus obtain substantial performance improvement compared to our previous work SERA and the existing stream data mining algorithms. Simulation results validate the effectiveness of the proposed MuSeRA algorithm.

This paper investigates the learning of a wide class of single-hidden-layer feedforward neural networks (SLFNs) with two sets of adjustable parameters, i.e., the nonlinear parameters in the hidden nodes and the linear output weights. The main objective is to both speed up the convergence of second-order learning algorithms such as Levenberg-Marquardt (LM), as well as to improve the network performance. This is achieved here by reducing the dimension of the solution space and by introducing a new Jacobian matrix. Unlike conventional supervised learning methods which optimize these two sets of parameters simultaneously, the linear output weights are first converted into dependent parameters, thereby removing the need for their explicit computation. Consequently, the neural network (NN) learning is performed over a solution space of reduced dimension. A new Jacobian matrix is then proposed for use with the popular second-order learning methods in order to achieve a more accurate approximation of the cost function. The efficacy of the proposed method is shown through an analysis of the computational complexity and by presenting simulation results from four different examples.

This paper investigates the selection of fuzzy rules for fuzzy neural networks. The main objective is to effectively and efficiently select the rules and to optimize the associated parameters simultaneously. This is achieved by the proposal of a fast forward rule selection algorithm (FRSA), where the rules are selected one by one and a residual matrix is recursively updated in calculating the contribution of rules. Simulation results show that, the proposed algorithm can achieve faster selection of fuzzy rules in comparison with conventional orthogonal least squares algorithm, and better network performance than the widely used error reduction ratio method (ERR).

In this letter, a standard postnonlinear blind source separation algorithm is proposed, based on the MISEP method, which is widely used in linear and nonlinear independent component analysis. To best suit a wide class of postnonlinear mixtures, we adapt the MISEP method to incorporate a priori information of the mixtures. In particular, a group of three-layered perceptrons and a linear network are used as the unmixing system to separate sources in the postnonlinear mixtures, and another group of three-layered perceptron is used as the auxiliary network. The learning algorithm for the unmixing system is then obtained by maximizing the output entropy of the auxiliary network. The proposed method is applied to postnonlinear blind source separation of both simulation signals and real speech signals, and the experimental results demonstrate its effectiveness and efficiency in comparison with existing methods.

A novel methodology is proposed for the development of neural network models for complex engineering systems exhibiting nonlinearity. This method performs neural network modeling by first establishing some fundamental nonlinear functions from a priori engineering knowledge, which are then constructed and coded into appropriate chromosome representations. Given a suitable fitness function, using evolutionary approaches such as genetic algorithms, a population of chromosomes evolves for a certain number of generations to finally produce a neural network model best fitting the system data. The objective is to improve the transparency of the neural networks, i.e. to produce physically meaningful "white box" neural network model with better generalization performance. In this paper, the problem formulation, the neural network configuration, and the associated optimization software are discussed in detail. This methodology is then applied to a practical real-world system to illustrate its effectiveness.

This paper proposes a novel hybrid forward algorithm (HFA) for the construction of radial basis function (RBF) neural networks with tunable nodes. The main objective is to efficiently and effectively produce a parsimonious RBF neural network that generalizes well. In this study, it is achieved through simultaneous network structure determination and parameter optimization on the continuous parameter space. This is a mixed integer hard problem and the proposed HFA tackles this problem using an integrated analytic framework, leading to significantly improved network performance and reduced memory usage for the network construction. The computational complexity analysis confirms the efficiency of the proposed algorithm, and the simulation results demonstrate its effectiveness.