摘要
提出了一种基于Logistic模型的惯性权重非线性调整策略,采用OpenMP多线程编程,在微机上实现了微粒群算法的多核并行计算。通过对BenchMark测试函数集中的5个函数进行测试,试验结果表明,采用基于Logistic模型的惯性权重非线性调整策略在算法成功率和收敛代数都优于线性调整策略,而基于OpenMP的微粒群多核并行计算使得计算速度得到提高。
A nonlinear adjustment strategy for inertia weight which is based on logistic model is proposed,and multi-core parallel computation of particle swarm optimization algorithm is realized which uses OpenMP multithread programming.Five function of BenchMark function set is tested.The results show that success rates and convergence times of algorithm which uses nonlinear adjustment strategy are superior to linear adjustment strategy.The calculation speed is improved which is based on OpenMP multi-core parallel computation of particle swarm optimization algorithm.
出处
《计算机工程与应用》
CSCD
北大核心
2010年第13期34-36,共3页
Computer Engineering and Applications
基金
山东省自然科学基金No.Y2007F25
中国石油大学优秀博士学位论文培育基金(No.B2007-05)~~
关键词
OPENMP
微粒群优化算法
多核并行计算
OpenMP
particle swarm optimization algorithm
multi-core parallel computation
