Novel bi-subgroup evolutionary programming based on chaotic mutation

Premature convergence is the fatal shortcoming of traditional evolutionary programming. In this paper, based on the analysis of traditional evolutionary programming premature convergence, a novel chaotic mutation based bi-subgroup evolutionary programming (CMBEP) algorithm is proposed. In the algorithm, evolutions of subgroups are paralleled performed with different mutation strategy. One of the subgroups takes chaotic mutation operator to explore the solution space separately, and the other subgroup searches the local part detailedly using exponential decrescent operator. Individuals, together with information, are exchanged while the population is reorganized. The simulations based on benchmarks confirm that CMBEP algorithm is better than classic evolutionary programming algorithm in the aspects of global optimization, convergence speed and the robustness.