Abstract: In order to solve the problems of optimization efficiency and inferior local search of cloud particle swarm algorithm which is based on cloud digital features (Ex, En, He) , this paper proposes two improvements, combines local search with global search which is based on the solution space transform, and achieves the evolution of the learning process and the variation operation according to the normal cloud particle. The improved algorithm is used to the function extreme optimization of muti-variables. Simulation results show that the improved algorithm has lower generation,higher efficiency, diverse population, and it proves that the improvement is efficient.

Key words: Cloud model, Optimization algorithm, Cloud particle swarm algorithm, Function optimization