A comparative study on differential evolution with other heuristic methods for continuous optimization

In this paper, we describe an optimization method based on differential evolution (DE). It shows good convergence properties with few parameters. However, the appropriate selection of the parameters is a difficult task. Hence, we here analyze the performance indexes of the DE algorithm to set the control parameters. Moreover, to identify the best parameter intervals, the DE approach is first compared to two different Particle Swarm Optimization (PSO) algorithms and then to a recent adaptive genetic algorithm (DABGA). The optimization of benchmark functions shows that the DE algorithm performs better than PSO and DABGA methods.