摘要
在继承综合学习粒子群算法(Comprehensive Learning Particle Swarm Optimizer,CLPSO)全局探索优势的基础上,引入具有高效收敛性能的传统局部搜索(Orthodox Local Search,OLS)方法,提出了基于拟熵自适应启动局部搜索策略的混合粒子群算法(Hybrid Particle Swarm Optimization algorithm with Adaptive starting strategy of Local Search based on Quasi-Entropy,ALSQE-HPSO).采用拟熵指标解决何时启动OLS这一关键问题.对8个标准函数的10维和20维问题的测试结果,表明了ALSQE-HPSO算法的性能优势.本文提出的算法也与包含两种基于CLPSO的改进算法和一种带OLS的粒子群算法在内的其他6种改进粒子群算法进行了对比,实验结果表明ALSQE-HPSO算法的性能优于对比算法.
Based on inheriting the advantage of global exploration of Comprehensive Learning Particle Swarm Optimizer( CLPSO), the Orthodox Local Search( OLS) approaches with efficient convergence are introduced and a Hybrid Particle Swarm Optimization algorithm with Adaptive starting strategy of Local Search based on Quasi-Entropy( ALSQE-HPSO)is proposed. A quasi-entropy index is utilized to solve the key issue of when to start OLS. The test results of 10-dimension and 20-dimension of eight benchmark functions show the performance advantages of the ALSQE-HPSO algorithm. The comparisons between the proposed algorithm and six other improved PSO algorithms, including two improved CLPSO algorithms and one PSO algorithm with OLS, are also made. The numerical results indicate that the performance of the ALSQEHPSO is superior to the compared algorithms.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2018年第1期110-117,共8页
Acta Electronica Sinica
基金
国家自然科学基金(No.61571336
No.61603280
No.71372202)
关键词
进化算法
粒子群优化
自适应策略
局部搜索
种群多样性
evolutionary algorithm
particle swarm optimization
adaptive strategy
local search
population diversity