摘要：To improve particle swarm optimization (PSO) computing performance, the centroid of particle swarm is firstly introduced in standard PSO model to enhance inter-particle cooperation and information sharing capabilities, then combining randomness and ergodicity of the strong chaotic motion and fast convergence of the simplex method, a novel particle swarm optimization algorithm with adaptive space mutation (CSM-CPSO) is proposed to improve local optimum efficiency and global convergence performance of PSO algorithm. Results of Benchmark function simulation and the material balance computation (MBC) in alumina production show the new algorithm has not only steady convergence and better stability, but also higher precision and faster convergence speed, and also can avoid the premature convergence problem effectively.
