Abstract:

As for the drop of particle diversity and the slow convergent speed of particle in the late evolution period when particle swarm optimization (PSO) is applied to solve high-dimensional multi-modal functions, a hybrid optimization algorithm based on the cat mapping, the cloud model and PSO is proposed. While the PSO algorithm evolves a certain of generations, this algorithm applies the cat mapping to implement global disturbance of the poorer individuals, and employs the cloud model to execute local search of the better individuals; accordingly, the obtained best individuals form a new swarm. For this new swarm, the evolution operation is maintained with the PSO algorithm, using the parameter of pop distr to balance the global and local search capacity of the algorithm, as well as, adopting the parameter of mix gen to control mixing times of the algorithm. The comparative analysis is carried out on the basis of 4 functions and other algorithms. It indicates that this algorithm shows faster convergent speed and better solving precision for solving functions particularly those high-dimensional multi-modal functions. Finally, the suggested values are proposed for parameters pop distr and mix gen applied to different dimension functions via the comparative analysis of parameters.