Modification And Application Of Particle Swarm Optimization Algorithm

Particle swarm optimization algorithm(PSOA) is an evolutionary computation technique based on swarm intelligence optimization algorithm,which was proposed by Kennedy and Eberhart in 1995. Its main feature is the principle of simple, less parameters, easy to converge and implement. Therefore, PSOA attractes the majority of experts and scholars'attention, and gradually becomes a new hotspot once it is proposed. However, there are still some deficiencies in PSOA, which include the slowly converging speed of PSOA in later period, premature convergence and easy to fall into local optimal solution.This paper does some research on the modifications and applications of PSOA.The detailed jobs are as follows:(1) In this paper, the convergence of simplified model of the algorithm was analyzed. Using the knowledge of matrix theory, we obtained a certain relationship between inertia weight and acceleration coefficient which makes the algorithm convergent, and convergence domain was given.(2) An improved inertia weight strategy was proposed. It obtains the strategy of dimensional weight calculation by using vector operation to analyze to the evolutionary formula of particle, accelerates the convergent speed and improves the global searching ability.(3) An improved particle swarm optimization algorithm was proposed. The algorithm uses a piecewise inertia weight strategy,which uses aforementioned weight strategy to evolve in the early and a adaptive weight strategy irrelevant to dimension in the later, as each dimensional difference decrease with the particles'evolving. Besides, in order to increase the diversity of the population, the chaos mutation strategy was used. The performance testing for six benchmark functions shows that the algorithm is an effective and improved algorithm.(4) A hybrid particle swarm optimization algorithm for solving nonlinear equations was devised, which combined the improved particle swarm optimization algorithm with Monte Carlo method. In order to improve searching accuracy, the Monte Carlo was used to search further after the global search of the improved particle swarm algorithm. The experimental results show that the algorithm is a highly practical algorithm.Finally,the whole research on texts were summarized,and further research directions were indicated.