A novel differential evolution algorithm with a self-adaptation parameter control method by differential evolution

Differential evolution (DE) is a simple yet powerful smart computing technique for numerical optimization. However, the performance of DE significantly relies on its parameters (scale factor F and crossover rate CR) of trial vector generating strategy. To address this issue, we propose a new DE variant by introducing a new parameter self-adaptation method into DE, called ADEDE. In ADEDE, a parameter population is established for the solution population, which is also updated from generation to generation based on the differential evolution under the basic principle that the good parameter individuals will go into the next generation at a high probability, while the bad parameter individuals will be updated by learning from the good parameter individuals at a large probability. To validate the efficiency of the proposed parameter self-adaptation method, the comparison experiments are tested on 22 benchmark functions. The experimental results show that the performance of classical DE can be significantly improved by our parameter self-adaptation method, and our method is better than or at least comparable to some other parameter control techniques.