Improved Differential Evolution for Function Optimization
Zhigang Zhou · 2010
This paper presents an improved differential evolution (DE) algorithm to enhance the performance of DE. The proposed approach is called MPTDE which employs a novel mutation operator. The main idea of MPTDE is to conduct a mutation on each individual and select a fitter one between the current one and the mutated one as the new current individual. In order to verify the performance of MPTDE, we test it on ten well-known benchmark functions. The experimental results show that MPTDE outperforms DE on majority of test functions.