Dual Mutation Strategies and Dual Crossover Strategies for Differential Evolution
Sheng-Ta Hsieh, Huang-Lyu Wu, Tse Su
Abstract
Sheng-Ta Hsieh, Huang-Lyu Wu, Tse Su
Abstract
In this paper, there are two mutation strategies and two crossover strategies are involved for enhancing solution searching ability of Differential Evolution (DE). These strategies will be activated according to current solution searching status. The elitist mutation will guide particles toward to solution space around the elitist particles, and the random to real-rand mutation can prevent particles form fall into local optimum. Both elitist crossover and one-cut-point crossover can produce potential particles for deeply search the basin of solution space. In the experiments, 25 test functions of CEC 2005 are adopted for testing performance of proposed method and compare it with 4 DE variants. From the results, it can be observed that the proposed method exhibits better than related works for solving most test functions.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In this paper, there are two mutation strategies and two crossover strategies are involved for enhancing solution searching ability of Differential Evolution (DE). These strategies will be activated according to current solution searching status. The elitist mutation will guide particles toward to solution space around the elitist particles, and the random to real-rand mutation can prevent particles form fall into local optimum. Both elitist crossover and one-cut-point crossover can produce potential particles for deeply search the basin of solution space. In the experiments, 25 test functions of CEC 2005 are adopted for testing performance of proposed method and compare it with 4 DE variants. From the results, it can be observed that the proposed method exhibits better than related works for solving most test functions.
Key concepts: Crossover, Differential evolution, Mutation, Dual (grammatical number), Mathematical optimization, Computer science, Mathematics, Algorithm