Two-layer particle swarm optimization for unconstrained optimization problems
In this article, a two-layer particle swarm optimization (TLPSO) is proposed to increase the diversity of the particles so that the drawback of trapping in a local optimum is avoided. In order to design the TLPSO, a structure with two layers (top layer and bottom layer) is proposed so that M swarms of particles and one swarm of particles are generated in the bottom layer and the top layer, respectively. Each global best position in each swarm of the bottom layer is set to be the position of the particle in the swarm of the top layer. Therefore, the global best position in the swarm of the top layer influences indirectly the particles of each swarm in the bottom layer so that the diversity of the particles increases to avoid trapping into a local optimum. Besides, a mutation operation is added into the particles of each swarm in the bottom layer so that the particles leap the local optimum to find the global optimum. Finally, some optimization problems of different types of high dimensional functions are used to illustrate the efficiency of the proposed method.