Particle swarm optimization algorithm is relatively distinctive algorithm since the birth of swarm intelligence optimization algorithm, it is a kind of based on population search strategy of adaptive random search algorithm, because it is more simple and less parameters than any other swarm intelligent algorithm, so the algorithm is easier to implement. Then, the algorithm has been widely used in neural network, support vector machine, data mining, engineering application and even in the areas of biological chemistry by the attention of many researchers at home and abroad, since it was put forward.Firstly, this paper analyzed the background and significance of the particle swarm algorithm, and learned more about the knowledge of the development process of particle swarm optimization algorithm, and found the shortcoming of particle swarm algorithm.Secondly, to deal with the problem of premature convergence and low precision of the standard particle swarm optimization algorithm, a particle swarm optimization algorithm based on two-population search and self-adaption search of each particle, is proposed. Every particles adjust their search speed based on their fitness, for adapting their search state. Meanwhile, the algorithm divide the population into two sub-population to coordinate the search state, and use Tent chaos model to search locally precisely. The proposed algorithm had been applied to optimize seven typical functions and compared with other swarm optimization algorithms that had been proposed for experimenting the performance.Finally, the proposed algorithm has been used to deal with the problems of the choice of the logistics distribution center and the parameters optimization of support vector machine. And the results show that the proposed method can effectively speed up the convergence and improve the stability. |