摘要
针对粒子群优化算法在寻优时容易出现早熟现象,提出在粒子群收敛停滞时,从种群中随机选择粒子进行共轭梯度法计算,通过引入共轭梯度算法计算的信息来影响粒子速度的更新,以保持群体的活性,从而打破群体信息陷入局部最优的状况.不同于传统的粒子群算法,该算法有机地结合了粒子群的全局搜索能力和共轭梯度法的强大局部搜索能力,从而在一定程度上有效地克服了粒子群早熟的缺点.仿真计算结果表明,该改进粒子群的方法对于不同维数的非线性函数具有很好的寻优效果.
To prevent the problem of premature convergence frequently appeared in the particle swarm optimization （PSO）, a method is proposed,which selects particles stochastically to perform the conjugate gradient algorithm when the PSO stagnates. The calculation information of the conjugate gradient algorithm is employed to affect the update of the particle speed so as to maintain the particle activation and avoid local optima. Unlike the existing PSO algorithms, the presented method integrates the global search ability of the PSO and the powerful local search ability of the conjugate gradient algorithm. Thus, the problem of premature convergence of the PSO algorithm is prevented. Simulation results show that the method has better performance for different dimensioned nonlinear functions.
出处
《控制与决策》
EI
CSCD
北大核心
2008年第5期511-514,519,共5页
Control and Decision
基金
国家杰出青年科学基金项目(60425310)
关键词
粒子群优化
共轭梯度法
全局最优
Particle swarm optimization
Conjugate gradient algorithm
Global optimum
