Research On Modification Of Particle Swarm Optimization Algorithm Based On Control Methods

Particle swarm optimization algorithm is one kind of swarm algorithm with profundity intelligence background, which has characters of can be applied in a flexible way, achieved easily and cooperative search. Theses characters make it become an effective way to solve complex optimization problems. Therefore, analysis evolution characters and evolution rules of particle swarm optimization algorithm is a subject which not only has both theory and application significance but also has significant practical significance for expand the area of algorithm application.This article based on control method; discussed particle swarm optimization algorithm convergence, structure amelioration and application. The specific research content was as follows:(1) Brief introduction of basic particle swarm optimization algorithm. This article expounded fundamental principle and actualize flow of particle swarm optimization algorithm, systematically summarize theory research and application research status quo of algorithm from different aspects. It also gave a synopsis introduction of some unsolved problems of particle swarm optimization algorithm research.(2) Analyzed convergence of basic particle swarm optimization algorithm. This article based on one kind of classical particle trajectory convergence analysis method, used mathematic deduction and control system stability criterion and discussed single particle trajectory convergence under two different situations. One was Pbest Time-varying and Gbest Time-invariant, and the other one was Pbest and Gbest Time-varying. After these analyses it obtained some parameter constraints which had more common sense to ensure algorithm. A verdict of simulation validated the efficiency and rationality of parameter constraints of this article.(3) Analyzed fuzzy PID-PSO algorithm and convergence. Based on particle swarm optimization algorithm evolution equation model, algorithm could be seen as a double input single output feedback control system. According to this analysis, this article advanced fuzzy PID-PSO algorithm and made some systematic analysis of algorithm convergence. A verdict of simulation indicated, fuzzy PID-PSO algorithm has good performance at optimization precision and efficiency on condition that convergence. Fuzzy PID-PSO balanced local exploitation and global exploration ability effectively. In Anderson system used fuzzy PID-PSO algorithm to optimize parameter of power system stabilizer (PSS). Eigen value and damping coefficient analysis indicated that PSS parameter which used fuzzy PID-PSO to optimize improve the stability of system.(4) Based on stochastic Robust Control particle swarm optimization algorithm, limited convergence control of PSO algorithm could be seen as Robust Control problem. This article begun with analyzing evolution equation of basic particle swarm optimization algorithm, then translated it to one kind of stochastic Discrete System difference model. Through brought in control variable and used the robust control method to achieve algorithm limited convergence. A verdict of simulation indicated, under robust controlled condition and control rule, basic PSO algorithm could convergence within lesser evolution generations. Moreover, simulation conclusion also proved the analyzed result of Robust controlled condition and control rule were efficiency.