An Improved Differential Evolution Algorithm Using Adaptive Multiple Mutation Strategies for the Economic Load Dispatch Problems

Abstract

An improved differential evolution algorithm using adaptive multiple mutation strategies (IMMSDE) is proposed to solve the economic load dispatch (ELD) problems with or without valvepoint effects (VPE). Unlike classical differential evolution (DE) algorithm, three mutation strategies and five adaptive parameters participate in the IMMSDE. A mutation strategy is randomly selected from the strategy pool in order to enlarge the search range, which is beneficial for preventing the solutions from falling into location optima. On the other hand, the adaptive parameters with learning ability improve the search accuracy and the speed of convergence. Additionally, a repair method is used to handle equality constraints, which enables IMMSDE to find feasible solutions rapidly. Four ELD cases are selected to testify the effectiveness of four DE algorithms on solving ELD problems. Results show that IMMSDE performs better than the other algorithms for the ELD problems and can always find the solutions satisfying the equality constraints.