Large-scale Optimization Using Center-based Differential Evolution with Dynamic Mutation Scheme
Hanan Hiba, Amin Ibrahim, Shahryar Rahnamayan · 2019
For several years the Differential Evolution (DE) algorithm has been an effective method for solving complex real-world optimization problems. However, when it comes to solving large-scale problems its performance deteriorates. In this paper, we propose five different dynamic center-based DE mutation schemes (DCDE) to solve large-scale optimization problems. In each generation, the proposed dynamic centerbased mutation strategies linearly divide the population into two different groups. Then, the first sub-population group utilizes center-based mutation scheme and the second sub-population employs the classical DE mutation. The proposed dynamic schemes are benchmarked on CEC 2013 large-scale optimization problems. The experimental results show that the overall performance of the proposed dynamic center-based mutation schemes better than the compared algorithms in solving LSGO problems.