摘要
基本粒子群算法在求解多数非线性函数优化问题时容易陷入局部极小,而陷入局部极小会导致搜索失败,在很大程度上限制了它的搜索能力,为解决此问题,提出改进粒子群算法,介绍了该算法的关键技术和具体步骤。改进粒子群算法分别采用混沌扰动机制、自反向机制及在迭代过程中重新初始适应值最差粒子等策略,用以解决局部最优及增强算法的种群多样性。还对改进算法进行了评估验证和仿真实验,实验结果证明,改进算法在搜索能力上有明显提高,能够较好地解决复杂优化问题。
To the problem that the basic particle swarm algorithm could easily plunge into the local minimum and cause low success search rate in case of solving nonlinear function optimization, an adaptive particle swarm algorithm is proposed. The key technologies and the specific steps of the algorithm are introduced. The improved particle swarm algorithm uses strategies such as chaotic mechanism, self-reverse mechanism and re-initializing the particle with worst adapting values in the iterative process. In this way, the problem of basic particle swarm algorithm could be solved and population diversity could be enhanced. The result of verification and simulation shows that the improved particle swarm algorithm could significantly improve the optimization efficiency and effectiveness in solving large-scale complex optimization problems.
出处
《控制工程》
CSCD
北大核心
2011年第6期841-844,909,共5页
Control Engineering of China
基金
"十一五"国家科技支撑计划(2006BAE03A06)
关键词
粒子群优化算法
混沌
自适应
末位重置
particle swarm optimization algorithm
Chaos
adaptive
last-initialized