Hybrid Particle Swarm Optimization And Its Applications

pattern search method and particle swarm optimization are combined to formHooke-Jeeves-PSO to solve nonlinear equations for the shortcomings of PSO search precision isnot high, the effect of HJ-PSO is better than PSO, social cognitive algorithms and the proposedQuasi-PSO algorithm; inertia weight factor is improved for the shortcomings of the slowconvergence rate of PSO, and the Nelder-Mead simplex method is inserted into PSO, theimproved algorithm is better than the residual method, simplex method and the PSO, etc.; thePSO-ACO algorithm is difficult to solve the TSP problems large-scale, GA, PSO and ACO aremixed into a novel algorithm with the crossover strategy in this paper ,it is used to solve threekinds scales of large-,middle-,small- on TSP problems, the effect is not only excellent for DPSO,ACO and PSO-ACO algorithm but also superior to six different ant colony algorithms;immunealgorithm and particle swarm optimization is combined and designed a formula of self-adaptivepenalty for solving constrained optimization problems for the bottleneck of the PSO algorithm insolving constrained optimization problems, the proposed algorithm is by 13 benchmachfunctions and three engineering examples to illustrate the performance of adaptive immuneparticle swarm optimization algorithm is equivalent to hybrid particle swarm algorithm based thefeasibility of rule, the improved algorithm is better than not only the immune algorithm and thePSO algorithm but also better than the co-evolutionary algorithms and the other four algorithms;the paper design a new pareto-PSO algorithm and it is used to solve the multi-objective andmulti-constrained design problem on mechanical parts because standard particle swarmoptimization algorithm can not solve it, it is better than multi-objective particle swarmoptimization algorithm based on weight method and multi-objective genetic algorithm basedon Gray Fuzzy decision making and gray theory-based particle swarm optimization.