Differential Evolution with Grey-based Hybrid Trial Vector Generation Strategies

： Ming-Feng Yeh, Min-Shyang Leu, Ti-Hung Chen and Po-Jung Huang
： Control parameters， Crossover strategy， Differential evolution， Grey relational analysis， Mutation strategy.

： This study attempts to propose a novel grey-basedtrial vector generation strategies to improve the searchperformance of differential evolution (DE) algorithm.The proposed strategies involve two main components.One is our previous work, differential evolution withgrey-based adaptive mutation factor (DE_GAMF), andthe other is the hybrid crossover strategy. The hybridcrossover strategy consists of a pool of distinct crossoverstrategies and randomly selects one to generate the trialvector for the associated target vector. DE withgrey-based hybrid trial vector generation strategies arealso applied to solve the optimization problems of eightbenchmark functions for illustration. Simulation resultsshow that the proposed approach could perform wellsearch performance on most of the test functions.

Journal of Grey System(Volume 16)
： 251~258
： 2013