| Summary: | 碩士 === 聖約翰科技大學 === 電機工程系碩士班 === 97 === This paper presents a simulated annealing liked particle swarm optimization algorithm ,with simulated annealing and the particle swarm optimization. The modification of both the simulated annealing and particle swarm optimization intends to more accurate convergence and produce faster.In this paper, a fixed number of 200,000 function evaluation were used as the termination criterion for all of the algorithms.
First of all, the sensitivity analysis of parameters for four algorithms, and then for four algorithms were tested and 20 were the best of the different dimensions of function tests, as well as simulated annealing liked particle swarm optimization excellent law and particle swarm optimization convergence excellent method to do comparison.The results show that excellent simulated annealing liked particle swarm optimization algorithm, both in the efficiency of the search of solutions and convergence speed were superior to other three algorithms (simulated annealing, genetic algorithms and particle swarm optimization).
|
|---|
