Min-Shyang Leu

Grey particle swarm optimization

With the help of grey relational analysis, this study attempts to propose two grey-based parameter automation strategies for particle swarm optimization (PSO). One is for the inertia weight and the other is for the acceleration coefficients. By the proposed approaches, each particle has its own inertia weight and acceleration coefficients whose values are dependent upon the corresponding grey relational grade. Since the relational grade of a particle is varying over the iterations, those parameters are also time-varying. Even if in the same iteration, those parameters may differ for different particles. In addition, owing to grey relational analysis involving the information of population distribution, such parameter automation strategies make an attempt on the grey PSO to perform a global search over the search space with faster convergence speed. The proposed grey PSO is applied to solve the optimization problems of 12 unimodal and multimodal benchmark functions for illustration. Simulation results are compared with the adaptive PSO (APSO) and two well-known PSO variants, PSO with linearly varying inertia weight (PSO-LVIW) and PSO with time-varying acceleration coefficients (HPSO-TVAC), to demonstrate the search performance of the grey PSO.

Particle swarm optimization with grey evolutionary analysis

Based on grey relational analysis, this study attempts to propose a grey evolutionary analysis (GEA) to analyze the population distribution of particle swarm optimization (PSO) during the evolutionary process. Then two GEA-based parameter automation approaches are developed. One is for the inertia weight and the other is for the acceleration coefficients. With the help of the GEA technique, the proposed parameter automation approaches would enable the inertia weight and acceleration coefficients to adapt to the evolutionary state. Such parameter automation behaviour also makes an attempt on the GEA-based PSO to perform a global search over the search space with faster convergence speed. In addition, the proposed PSO is applied to solve the optimization problems of twelve unimodal and multimodal benchmark functions for illustration. Simulation results show that the proposed GEA-based PSO could outperform the adaptive PSO, the grey PSO, and two well-known PSO variants on most of the test functions.

Financial Distress Prediction Model via GreyART Network and Grey Model

This study attempts to use GreyART network and grey model to construct a financial distress prediction model. The inputs used to train the network are the historical data containing 17 different financial ratios of 22 healthy and 5 distressed Taiwan’s listed banks. With the help of the developed performance index, this study also proposes a growing extraction method for financial variables not only to further improve the classification ability in the training and testing phases, but also to use fewer extracted variables to build the financial distress prediction model. Simulation results show that the optimal condition is the one using four extracted variables as inputs and the vigilance threshold of 0.80. Under this condition, the proposed method generates only two clusters with corresponding classification hit rates of 96.30% and 95.24% for the training and testing results, respectively.

System identification using grey-based adaptive particle swarm optimization

A grey-based adaptive particle swarm optimization (PSO) applied to the system identification for a class of nonlinear systems is proposed in this study. In the grey-based adaptive PSO, the inertia weights are dynamically adapted according to the results of grey relational analysis at each generation. Every particle has its own inertia weight which may differ for different particles or different generations. Besides, this study attempts to utilize the grey-based adaptive PSO to identify an unknown system whose structure is assumed to be known in advance. Simulation results are compared with PSO with a linearly varying inertia weight and PSO with adaptive inertia weight to demonstrate the search performance of the grey-based adaptive PSO.

ART-type CMAC network classifier

In this paper, the learning process of ART 2 (adaptive resonant theory) network is applied to construct the structure of cerebellar model articulation controller (CMAC) to form an ART-type CMAC network. The proposed updating rule is in an unsupervised manner as the ART 2 network or the self-organizing map (SOM), and could equally distribute the learning information into the association memory locations as the CMAC network. If the winner fails a vigilance test, a new state is created; otherwise, the memory contents corresponding to the winner are updated according to the learning information. Like SOM, the proposed network also has a neighborhood region, but the neighborhood region is implicit in the network structure and need not be defined in advance. This paper also analyzes the convergence properties of the ART-type CMAC network. The proposed network is applied to solve data classification problems for illustration. Experiment results demonstrate the effectiveness and feasibility of the ART-type CMAC network in solving five benchmark datasets selected from the UCI repository.

Modified Gaussian barebones differential evolution with hybrid crossover strategy

Inspired by Gaussian barebones differential evolution (GBDE), this study attempts to propose a new Gaussian mutation strategy, termed by GBDE/best-rand, to improve the solution accuracy. This study also proposes a hybrid crossover strategy, the hybridization of the binomial and arithmetic crossover strategies, for differential evolution (DE) to further balance the global search ability and convergence rate. The GBDE with the proposed Gaussian mutation and hybrid crossover strategies is also almost parameter free. The search performance of the proposed modified GBDEs is compared with two standard DEs (DE/rand/1 and DE/best/1), the GBDE and its modified version in terms of solution accuracy, convergence speed and reliability. Simulation results demonstrate the effectiveness of the proposed modified GBDE algorithm.

Grey-Based particle swarm optimization algorithm

In order to apply grey relational analysis to the evolutionary process, a modified grey relational analysis is introduced in this study. Then, with the help of such a grey relational analysis, this study also proposed a grey-based particle swarm optimization algorithm in which both inertia weight and acceleration coefficients are varying over the generations. In each generation, every particle has its own algorithm parameters and those parameters may differ for different particles. The proposed PSO algorithm is applied to solve the optimization problems of twelve test functions for illustration. Simulation results are compared with the other three variants of PSO to demonstrate the search performance of the proposed algorithm.

Optimizations of PID controllers by grey-based particle swarm optimizations

Based on grey relational analysis, this paper attempts to propose a grey-based particle swarm optimization (PSO). The central idea of the proposed approach is that the determination of the algorithm parameters (the inertia weight and the acceleration coefficients) for a particle is depended upon the grey relational grade of that particle. The algorithm parameters are varying over the generations. Also they may differ from different particles. Besides, the grey-based PSO algorithm is also applied to optimize the parameters of the proportional-integral-derivative (PID) controller. Simulation and experiment results are compared with the standard PSO algorithm and genetic algorithm to demonstrate the search performance of the proposed method.

Tuning of PID controllers with CMAC-based genetic algorithm

Based on cerebellar model articulation controller (CMAC), this paper attempts to propose a new chromosome representation scheme for representing real number parameters in genetic algorithms (GAs), which is termed CMAC-based GAs. The central idea of CMAC-based GAs is that each memory unit in a CMAC network is regarded as a real-valued gene in GAs. By this way, a chromosome is represented as those memory units addressed by a specific state and each gene involves partial information about a potential solution, not a potential solution as the conventional GAs. In addition, the corresponding crossover and mutation operators are also derived for the proposed GAs. The proposed CMAC-based GAs is applied to optimize the parameters of the proportional-integral-derivative controller. Simulation results are compared with several previous findings to demonstrate the search performance of the proposed method.

System identification using differential evolution with mean-best mutation

This paper attempts to propose a new mutation strategy, termed the mean-best mutation strategy, for differential evolution (DE) algorithm to enhance global search ability and to avoid premature convergence. In the proposed mutation strategy (denoted by DE/mBest/1), the base vector is the mean of the p top-ranked individuals, and denoted by mBest. That is, the randomly selected base vector of DE/rand/1 or the best vector of DE/best/1 is replaced by mBest in the proposed scheme. DE/mBest/1 is applied to identify an unknown system whose structure is assumed to be known in advance. The search performance of DE/mBest/1 is compared with two standard DEs(DE/rand/l and DE/best/1), two of our previous works and 2-Opt based DE in terms of parameter accuracy, convergence speed and reliability. Simulation results demonstrate the effectiveness of the proposed DE algorithm.