Analysis and operation of three different forms probabilistic particle swarm optimization algorithm

3rd International Conference on Advances in Energy, Environment and Chemical Engineering

能源学;生态环境科学;化学工业

Sun, Tao^1,2 ; Xu, Minghai^1

College of Pipeline and Civil Engineering, China University of Petroleum, Qingdao, Shandong

266580, China^1

Shengli College, China University of Petroleum, Dongying, Shandong

257000, China^2

关键词: Calculation accuracy;    Global search ability;    Optimization problems;    Particle swarm optimization algorithm;    Position equations;    Quantum-behaved particle swarm algorithms;    Standard test functions;    Stochastic simulation methods;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/69/1/012158/pdf
DOI  :  10.1088/1755-1315/69/1/012158

学科分类：环境科学（综合）

来源: IOP

【 摘 要 】

Quantum-behaved Particle Swarm algorithm(QPSO) is a kind of probabilistic PSO algorithm based on quantum theory, which has been successfully applied to solve many optimization problems. This paper analyzes the conditions of the probability density function should satisfy in QPSO, and constructs three functions that meet the requirements: exponential form, normal form and power form; thus obtains three position equation of particle by using the stochastic simulation method; then compares the convergence of the three forms PSO algorithm, and uses different types of standard test functions to evaluate them. The results show that exponential form and power form PSO has better convergence speed and calculation accuracy than standard PSO. In three different forms algorithm, power form PSO has better global search ability, and more suitable for solving high-dimensional and Multi-extremum optimization problem.