Differential evolution with automatic parameter configuration for solving the CEC2013 competition on Real-Parameter Optimization

Abstract

The performance of Differential Evolution (DE) algorithms is known to be highly dependent on its search operators and control parameters. The selection of the parameter values is a tedious task. In this paper, a DE algorithm is proposed that configures the values of two parameters (amplification factor and crossover rate) automatically during its course of evolution. For this purpose, we considered a set of values as input for each of the parameters. The algorithm has been applied to solve a set of test problems introduced in IEEE CEC'2013 competition. The results of the test problems are compared with the known best solutions and the approach can be applied to other population based algorithms.