The Research Of Improving Particle Swarm Optimization Algorithm

Because the Intelligent Optimization Algorithm was widely used in the practical problems,now, research of improving intelligent algorithm has become the main direction of research.Compared with other intelligent optimization algorithms, Particle Swarm Optimization (PSO)not only has great advantages of simple principle, few parameters and fast convergence rates, butalso is easier to perform. Because of its advantages, PSO is favored by many scholars. To thefunction optimization problem, many scholars have proposed many improved PSO algorithms.The main contents of this paper are as follows:1. The Quadratic Interpolation Particle Swarm Optimization Algorithm. Because PSOalgorithm is easy to fall in local optima and is low accuracy in searching, a new PSO algorithm isproposed based on Quadratic Interpolation. Quadratic Interpolation and Standard Particle SwarmOptimization algorithm are combined in this algorithm. After the update of the particle velocityand position, two positions from set of the current personal best position are closed at random. Anew position is produced by the quadratic interpolation given through the current personal bestposition and the global best position positions, so the global searching performance isimproved.Simulation experimental results of some classic benchmark functions indicate that thenew algorithm greatly improves the searching efficiency and the convergence.2. To improve the local searching performance and diversity of evolution population of PSOalgorithm, a particle swarm optimization algorithm is proposed based on the basic local search.Firstly, Local search method is given and applied to the global best position. Then, when theglobal best position is not improved, chaos optimization strategy is applied in it. The proposedalgorithm ensures the ability of local search and global search.The algorithm is applied to sometest functions and the results show the efficiency and effectiveness of this method.