An improved differential evolution algorithm based on suboptimal solution mutation

In order to improve the drawbacks of DE algorithm with DE/best/1 such as the rapid convergence speed and local optimum, this paper proposes an improved DE algorithm. Based on the DE/best/1 mutation operator, a new mutation operator is constructed. The best M individuals are summed as a new individual to replace the base individual of the DE/best/1. This is helpful to avoid falling into local optimum for the fast convergence. Simulation experiments demonstrate that the proposed algorithm outperforms some standard DE variants.