In multiobjective evolutionary algorithms, how to select the optimal solutions from the offspring candidate set significantly affects the optimization process. Usually, the selection process is largely based on the real objective values or surrogate model estimating objective values. However, these selection processes are very time consuming sometimes, especially for some real optimization problems. Recently, some researches began to employ supervised classification to assist offspring selection, but these works are difficult to prepare the exact positive and negative samples or time consuming of parameter tuning problems. In order to solve these disadvantages, we propose a decomposition based multiobjective evolutionary algorithm with semi-supervised classification. This approach using random sampling and non-dominated sorting to construct semi supervised classifier. In each generation, a set of candidate solutions are generated for each subproblem and only good solutions are reserved by classifier. If there is more than one good solutions, we calculate each of good solutions by real objective function and choose the best one as the offspring solution. Based on the typical decomposition based multiobjective evolutionary algorithm MOEA/D, we design algorithm framework through integrating the novel offspring selection process based on semi-supervised classification. Experiments show that the proposed algorithm performs best in most test cases and improves the performance of MOEA/D.

Solving constrained multiobjective optimization problems is one of the most challenging areas in the evolutionary computation research community. To solve a constrained multiobjective optimization problem, an algorithm should tackle the objective functions and the constraints simultaneously. As a result, many constraint-handling techniques have been proposed. However, most of the existing constraint-handling techniques are developed to solve test instances (e.g., CTPs) with low dimension and large feasible region. On the other hand, experimental comparisons on different constraint-handling techniques remain scarce. In view of these two issues, in this paper we first construct 18 test instances, each of which exhibits different properties. Afterward, we choose three representative constraint-handling techniques and combine them with nondominated sorting genetic algorithm II to study the performance difference on various conditions. By the experimental studies, we point out the advantages and disadvantages of different constraint-handling techniques.

Differential Evolution (DE) is a popular paradigm of evolutionary algorithms, which has been successfully applied to solve different kinds of optimization problems. To design an effective DE, it is necessary to consider different requirements of the exploration and exploitation at different evolutionary stages. Motivated by this consideration, a new DE with a two-stage optimization mechanism, called TSDE, has been proposed in this paper. In TSDE, based on the number of fitness evaluations, the whole evolutionary process is divided into two stages, namely the former stage and the latter stage. TSDE focuses on improving the search ability in the former stage and emphasizes the convergence in the latter stage. Hence, different trial vector generation strategies have been utilized at different stages. TSDE has been tested on 25 benchmark test functions from IEEE CEC2005 and 30 benchmark test functions from IEEE CEC2014. The experimental results suggest that TSDE performs better than four other state-of-the-art DE variants.

Ordinal regression has a wide range of applications, while it is intractable to be solved when lacking sufficient labeled data. In this paper, we propose an evolutionary semi-supervised kernel Fisher discriminant approach for ordinal regression. The proposed algorithm obtains the projection and thresholds by incorporating the unlabeled data with a weighting scheme, where the weights indicate the degrees of contributions to the class distribution by different training instances. The projection maps the original data to a one-dimensional space, and the thresholds are used for the prediction. The weights are computed with a label propagation method first. However, it is not always accurate. In order to further tune the weights to be more accurate, the differential evolution algorithm is applied here in this work. By a delicate weight update rule, the weights can be evolved indirectly. This tuning scheme makes the size of evolutionary individual just associated with the number of ranks rather than the number of instances. The experimental studies demonstrate that our algorithm can effectively use unlabeled data and yield satisfactory learning performance.

Nonlinear equation systems may have multiple optimal solutions. The main task of solving nonlinear equation systems is to simultaneously locate these optimal solutions in a single run. When solving nonlinear equation systems by evolutionary algorithms, usually a nonlinear equation system should be transformed into a kind of optimization problem. At present, various transformation techniques have been proposed. This paper presents a simple and generic transformation technique based on multiobjective optimization for nonlinear equation systems. Unlike the previous work, our transformation technique transforms a nonlinear equation system into a biobjective optimization problem that can be decomposed into two parts. The advantages of our transformation technique are twofold: 1) all the optimal solutions of a nonlinear equation system are the Pareto optimal solutions of the transformed problem, which are mapped into diverse points in the objective space, and 2) multiobjective evolutionary algorithms can be directly applied to handle the transformed problem. In order to verify the effectiveness of our transformation technique, it has been integrated with nondominated sorting genetic algorithm II to solve nonlinear equation systems. The experimental results have demonstrated that, overall, our transformation technique outperforms another state-of-the-art multiobjective optimization based transformation technique and four single-objective optimization based approaches on a set of test instances. The influence of the types of Pareto front on the performance of our transformation technique has been investigated empirically. Moreover, the limitation of our transformation technique has also been identified and discussed in this paper.

The aim of designing a sensor scheduling scheme for target tracking in wireless sensor network is to improve the tracking accuracy, balance the network energy and prolong the network lifespan. It is viewed as a multi-objective optimization problem. A modified matrix-coded parallel genetic algorithm (MPGA) is proposed in which multiple subpopulations evolve synchronously and satify the specific constraint arised from the senario of multi-target tracking that a sensor can only track just one target. Simulation results show that MPGA, compared with traditional genetic algorithm, converges to the better result with higher speed when applied in multi-target tracking in wireless sensor network. And our proposed distributed sensor scheduling scheme based on MPGA outperforms than existed schemes.

Detection of arrow traffic light is a focal point research in autonomous vehicle, and in urban environment it is the basic technique. However, most researches mainly concern the circular traffic lights. A novel algorithm is proposed in this paper to resolve the problems of detection and recognition of arrow traffic lights. Two sub-modules, detection module and recognition module, are introduced in the main framework. In detection submodule, the color space conversion, binarization and morphology features filtering methods are performed to get the regions of candidates of blackboards. For getting the regions of arrow of traffic lights, segmentation based on the YCbCr color space is used in the cropping image, which is cropped from original image by the region of blackboard. In recognition sub-module, Gabor wavelet transform and 2D independent component analysis(2DICA) are used to extract traffic light candidate's features for features of the arrow traffic lights. A library for recognition has been built, and experimental results show that rate of recognition exceeds 91%.

During the past decade, solving constrained optimization problems with evolutionary algorithms has received considerable attention among researchers and practitioners. Cai and Wang's method (abbreviated as CW method) is a recent constrained optimization evolutionary algorithm proposed by the authors. However, its main shortcoming is that a trial-and-error process has to be used to choose suitable parameters. To overcome the above shortcoming, this paper proposes an improved version of the CW method, called CMODE, which combines multiobjective optimization with differential evolution to deal with constrained optimization problems. Like its predecessor CW, the comparison of individuals in CMODE is also based on multiobjective optimization. In CMODE, however, differential evolution serves as the search engine. In addition, a novel infeasible solution replacement mechanism based on multiobjective optimization is proposed, with the purpose of guiding the population toward promising solutions and the feasible region simultaneously. The performance of CMODE is evaluated on 24 benchmark test functions. It is shown empirically that CMODE is capable of producing highly competitive results compared with some other state-of-the-art approaches in the community of constrained evolutionary optimization.

Based on our previous work, this paper presents a dynamic hybrid framework, called DyHF, for solving constrained optimization problems. This framework consists of two major steps: global search model and local search model. In the global and local search models, differential evolution serves as the search engine, and Pareto dominance used in multiobjective optimization is employed to compare the individuals in the population. Unlike other existing methods, the above two steps are executed dynamically according to the feasibility proportion of the current population in this paper, with the purpose of reasonably distributing the computational resource for the global and local search during the evolution. The performance of DyHF is tested on 22 benchmark test functions. The experimental results clearly show that the overall performance of DyHF is highly competitive with that of a number of state-of-the-art approaches from the literature.

We propose a method to recognize Chinese traffic police gesture in complex scene for intelligent vehicle. The gesture recognition is made by integrated nonparametric background modeling with human pose estimation. Firstly, dark channel prior and kernel density estimation are used to extract police's torso and arms from complex traffic environment as foreground region. Then, the coordinates of pixels in the upper and lower arms are determined by using max-covering scheme, which is based on a key observation that body part tiles maximally cover the foreground region and satisfy body plan. Finally, some typical police gestures can be recognized by rotation joint angle. The experimental results show that this method can obtain favorable results on a number of gesture sequences of traffic police.

This paper proposes a (μ + λ)-differential evolution and an improved adaptive trade-off model for solving constrained optimization problems. The proposed (μ + λ)-differential evolution adopts three mutation strategies (i.e., rand/1 strategy, current-to-best/1 strategy, and rand/2 strategy) and binomial crossover to generate the offspring population. Moreover, the current-to-best/1 strategy has been improved in this paper to further enhance the global exploration ability by exploiting the feasibility proportion of the last population. Additionally, the improved adaptive trade-off model includes three main situations: the infeasible situation, the semi-feasible situation, and the feasible situation. In each situation, a constraint-handling mechanism is designed based on the characteristics of the current population. By combining the (μ + λ)-differential evolution with the improved adaptive trade-off model, a generic method named (μ + λ)-constrained differential evolution ((μ + λ)-CDE) is developed. The (μ + λ)-CDE is utilized to solve 24 well-known benchmark test functions provided for the special session on constrained real-parameter optimization of the 2006 IEEE Congress on Evolutionary Computation (CEC2006). Experimental results suggest that the (μ + λ)-CDE is very promising for constrained optimization, since it can reach the best known solutions for 23 test functions and is able to successfully solve 21 test functions in all runs. Moreover, in this paper, a self-adaptive version of (μ + λ)-CDE is proposed which is the most competitive algorithm so far among the CEC2006 entries.

Trial vector generation strategies and control parameters have a significant influence on the performance of differential evolution (DE). This paper studies whether the performance of DE can be improved by combining several effective trial vector generation strategies with some suitable control parameter settings. A novel method, called composite DE (CoDE), has been proposed in this paper. This method uses three trial vector generation strategies and three control parameter settings. It randomly combines them to generate trial vectors. CoDE has been tested on all the CEC2005 contest test instances. Experimental results show that CoDE is very competitive.

Based on our previous work, a hybrid differential evolution is proposed in this paper and combined with an adaptive trade-off model to solve constrained optimization problems. The proposed hybrid differential evolution adopts three mutation strategies to generate mutant vectors. Unlike the existing methods, direction information has been utilized in this paper. The proposed method is tested on thirteen well-known benchmark test functions. Our method outperforms or performs similarly to two state-of-the-art approaches in terms of effectiveness and efficiency.

In this paper, we propose a context-based inverse quantization and show its application in wavelet image compression. The proposed method breaks the traditional one-to-one mapping of the quantization index to reconstruction value in inverse quantization while maps an index to several different reconstruction values according to the corresponding contexts of the index. By accurate context modeling, this method can reduce the quantization distortion significantly. Since the quantization indices used for encoding is not changed, this method does not increase the encoding bit rates except the negligible overhead.

This letter introduces a simple yet efficient rate control method, which is designed for lossless mode of JPEG2000. This method uses a fractional upward shift of bitplane to simulate more precise subband weighting. Compared to the previous integer upward shift of the bitplane, it achieves superior performance at most bit-rates without the cost of increased computational complexity.

In this paper, an adaptive tradeoff model (ATM) is proposed for constrained evolutionary optimization. In this model, three main issues are considered: (1) the evaluation of infeasible solutions when the population contains only infeasible individuals; (2) balancing feasible and infeasible solutions when the population consists of a combination of feasible and infeasible individuals; and (3) the selection of feasible solutions when the population is composed of feasible individuals only. These issues are addressed in this paper by designing different tradeoff schemes during different stages of a search process to obtain an appropriate tradeoff between objective function and constraint violations. In addition, a simple evolutionary strategy (ES) is used as the search engine. By integrating ATM with ES, a generic constrained optimization evolutionary algorithm (ATMES) is derived. The new method is tested on 13 well-known benchmark test functions, and the empirical results suggest that it outperforms or performs similarly to other state-of-the-art techniques referred to in this paper in terms of the quality of the resulting solutions.

A modified embedded trellis coded quantization (TCQ) for JPEG2000 is presented in this paper. The method for approximately inverting TCQ in the absence of the least significant bits is improved. Experimental results, presented using the optimal rate control algorithm and different embedded TCQ formulations, show that modified embedded TCQ yields significant performance improvement compared to the original one in JPEG2000.

Solving constrained optimization problems (COPs) via evolutionary algorithms (EAs) has attracted much attention recently. A new constrained optimization evolutionary algorithm by using good point set (COEAGP) is presented in this paper. In the process of population evolution, multi-objective optimization techniques and good point set in number theory are integrated into our algorithm. The approach transforms COP into a bi-objective optimization problem firstly. Then the crossover operator is designed by using the principle of good point set The purpose of the new crossover is to enrich the exploration and exploitation abilities of the approach proposed. The new crossover operator can produce a small but representative set of points as the potential offspring. After that the BGA mutation operator is applied to potential offspring for enhancing the diversity of the potential offspring population. Furthermore, the update operator incorporates Pareto dominance and the tournament selection operator to choose the best individuals in the current offspring for the next generation. The new approach is tested on 8 well-known benchmark functions, and the empirical evidence suggests that it is robust and efficient when handling linear/nonlinear equality/inequality constraints and that COEAGP outperforms or performs similarly to the other techniques referred in this paper in terms of the quality of the resulting solutions.

Good Nodes Set(GNS) is a concept in number theory. To overcome the deficiency of orthogonal design to handle constrained optimization problems(COPs), this paper presents a method that incorporate GNS principle to enhance the crossover operator of the evolution strategy (ES) can make the resulting evolutionary algorithm more robust and statically sound. In order to gain the rapid and stable rate of converging to the feasible region, traditional crossover operator is split into two steps. GNS initialization methods is applied to ensure the initial population span evenly in relatively large search space and reliably locate the good points for further exploration in subsequent iterations. The proposed method achieves the same sound results just as the orthogonal method does, but its precision is not confined by the dimension of the space. The simplex selected and diversity mechanism similar to Carlos's SMES is used to enrich the exploration and exploitation abilities of the approach proposed. Experiment results on a set of benchmark problems show the efficiency of our methods.

This paper presents a novel evolutionary algorithm (EA) for constrained optimization problems, i.e., the hybrid constrained optimization EA (HCOEA). This algorithm effectively combines multiobjective optimization with global and local search models. In performing the global search, a niching genetic algorithm based on tournament selection is proposed. Also, HCOEA has adopted a parallel local search operator that implements a clustering partition of the population and multiparent crossover to generate the offspring population. Then, nondominated individuals in the offspring population are used to replace the dominated individuals in the parent population. Meanwhile, the best infeasible individual replacement scheme is devised for the purpose of rapidly guiding the population toward the feasible region of the search space. During the evolutionary process, the global search model effectively promotes high population diversity, and the local search model remarkably accelerates the convergence speed. HCOEA is tested on 13 well-known benchmark functions, and the experimental results suggest that it is more robust and efficient than other state-of-the-art algorithms from the literature in terms of the selected performance metrics, such as the best, median, mean, and worst objective function values and the standard deviations.

Fault detection and diagnosis (FDD) is very important for wheeled mobile robots (WMRs). In this paper, an adaptive particle filter is developed to deal with unknown fault detection as well as known fault diagnosis for wheeled mobile robots. Two parameters are extracted from sample-based expression for a posteriori probability density: sum of unnormalized weight of samples, and Kullback-Leiber divergence of proposal distribution and posteriori distribution. Decision rules are derived to determine novel faults based on these parameters. Fault state space is adapted according the number of detecting novel fault. This method preserves the advantages of particle filter and can diagnose known faults as well as detect unknown faults. The method is testified on mobile robot fault diagnosis problem

A considerable number of constrained optimization evolutionary algorithms (COEAs) have been proposed due to increasing interest in solving constrained optimization problems (COPs) by evolutionary algorithms (EAs). In this paper, we first review existing COEAs. Then, a novel EA for constrained optimization is presented. In the process of population evolution, our algorithm is based on multiobjective optimization techniques, i.e., an individual in the parent population may be replaced if it is dominated by a nondominated individual in the offspring population. In addition, three models of a population-based algorithm-generator and an infeasible solution archiving and replacement mechanism are introduced. Furthermore, the simplex crossover is used as a recombination operator to enrich the exploration and exploitation abilities of the approach proposed. The new approach is tested on 13 well-known benchmark functions, and the empirical evidence suggests that it is robust, efficient, and generic when handling linear/nonlinear equality/inequality constraints. Compared with some other state-of-the-art algorithms, our algorithm remarkably outperforms them in terms of the best, mean, and worst objective function values and the standard deviations. It is noteworthy that our algorithm does not require the transformation of equality constraints into inequality constraints

This paper presents a novel evolutionary algorithm for constrained optimization. During the evolutionary process, our algorithm is based on multiobjective optimization techniques, i.e., an individual in the parent population may be replaced if it is dominated by a nondominated individual chosen from the offspring population. In addition, a model of population-based algorithm-generator and an infeasible solutions archiving and replacement mechanism are introduced. Furthermore, the simplex crossover is used as a recombination operator to enrich the exploration and exploitation abilities of the approach proposed. The new approach is tested on thirteen well-known benchmark functions, and the empirical evidences suggest that it is robust, efficient and generic when handling linear/nonlinear equality/inequality constraints. Compared with some other state-of-the-art algorithms, our algorithm remarkably outperforms them in terms of the best, median, mean, and worst objective function values and the standard deviations.

Fault detection and diagnosis (FDD) and fault tolerant control (FTC) are increasingly important for wheeled mobile robots (WMRs), especially those in unknown environments such as planetary exploration. Due to the importance of reliability and safe operation of WMRs, this paper presents a survey of state-of-the-art in FDD & FTC of WMRs under unknown environments. Firstly, we briefly introduce main components, typical kinematics models and fault models of WMRs and error models of inertial navigation sensors. Secondly, we discuss main approaches for FDD/FTC of WMRs, including multiple model based approach, particle filter based approach, sensor fusion based approach, layered fault tolerant architecture and so on. At last, the main challenges, difficulties and some future trends for the field are offered.

Fault detection and diagnosis (FDD) is increasingly important for wheeled mobile robots (WMRs). One of the most promising approaches is the so-called particle filter (also known as sequential Monte Carlo) method. In this paper, rule based inference and multiple particle filters are integrated to diagnose hard faults of WMR's dead reckoning system. The rule based inference method is employed to determine the states of the movement of the robot in plane and each state of movement is monitored with a particle filter. This approach presents a general framework to combine domain knowledge with particle filters. The key advantage of the proposed method is that it decreases the size of the state space for each particle filter. As a result, it decreases particle number and increases efficiency and accuracy for each particle filter. Experiment performed on a mobile robot shows the improvement in accuracy and efficiency.

This paper proposes a connectionist model to learn a spatial representation of the world based on temporal memory of perceptions and actions of the robot. It is constructed at run-time to merge the experiences and retrieved in later runs to guide the robot to perform the navigation task. A coding strategy is introduced to extract the directional information from the perception sequence, which endows the robot with localization ability. The Temporal Sequence Processing Network (TSPN) transforms routing knowledge learned from robot experiences into temporal characteristics of cell firing and enables the implicit building of a metric map. The navigation system integrating TSPN and a reactive safeguard module performs collision-free navigation, dynamic landmark and heading detection, route learning and path planning in a noisy world, which is tolerant of sensor inaccuracies and unexpected obstacles. The simulation and real world experiments demonstrate the flexibility and robustness of the system.

Some teaching works on the course of Artificial Intelligence (AI) at Purdue University given by Professor King-Sun Fu are introduced. Especially, some research works on expert systems at the AI course made by the graduate students and directed by Prof. Fu are dealt with. Three examples of expert systems, in quality control, robot programming, and product evaluation, are given respectively. In this way, the author want to cherish his emotion for Dr. K.S.Fu.

This paper uses the concepts and technologies of the expert systems to the robot planning. Some examples of these planning systems have been presented. Finally, the planning performances of these systems have been compared with the existing planning systems.