A Kind of Improved Uniform Particle Swarm Optimization Algorithm

Particle swarm optimization is usually random, which leads to random distribution of search quality and search speed. So the general improved particle swarm optimization is difficult to meet fast optimization needs of some actual engineering. Stocks in the key generation of PSO algorithm generated by uniform design method can make particles in the population maintain a better uniform distribution in the search space. However, due to restriction from the size of the equal level uniform table, the present uniform particle swarm algorithm is only applicable to the optimization calculation of particular population with identified size. For this lack, this paper presents four kinds of methods to generate key generation of the population, which will improve and enrich the uniform particle swarm algorithm, and it can meet the different size populations to be generated. At last, the performance of improved algorithm is tested and compared. Tests show that the improved algorithm not only can retain the original algorithm some advantages and the occurrence of particle aggregation and premature is effectively reduced, but also can apply to different population size.