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

As one kind of swarm intelligence optimization algorithms,particle swarm optimization (PSO) algorithms, with simple structure, less adjustable parameters,easy to implement and optimization ability characteristic, has become a research hotspot to solve complex optimization problems.Therefore,the analysis and research on PSO internal mechanism and laws of evolution and further improvement on the algorithm optimization ability has important theoretical and practical significance.Considering the randomness of the algorithm and taking advantage of the information generated in the optimization process,this thesis based on control thoughts carries out improvement research of particle swarm optimization from the perspective of process optimization. The specific research contents are as follows:this article firstly introduces fundamentals of particle swarm optimization algorithm by systematically reviewing theory research on the status quo and research results of algorithm at home and abroad. Because of the uncertainty caused by random factors inherently, part of the particle tracks are resulted in "oscillation" near the optimum value and the rapid convergence of the particle is impeded. This paper proposes a kind of particle swarm optimization algorithms based on optimal control for uncertain discrete-time systems.That is to say, this algorithm makes it become uncertain model and regards fast convergence as index by derivation PSO iterative formula based on basic particle optimization model. The simulation results indicate that by introducing the optimal controller the convergence rate improve obviously and the searching precision improves to some degree. At last, by analyzing the data information produced in the process of particle swarm optimization algorithm optimization such as the optimal value changes, diversity, average speed and so on, the PSO algorithm optimization process is defined as the diversity of feedback control, adaptation values optimal control and average speed feedback control process.Then, the idea of process optimization is introduced to the PSO algorithm based on the optimal value, diversity, average speed change information on the merits of the reflection process optimization and the optimal control decision rules is proposed after that. Consequently it could control the PSO algorithm optimization process from the optimal value, the diversity and average speed aspects. The simulation results demonstrate the feasibility and effectiveness of the algorithm.