摘要
针对基本粒子群算法容易陷入局部最优的缺点,将禁忌搜索算法中的禁忌思想与粒子群算法结合,提出了一种新的粒子群算法——禁忌粒子群算法(TPSO)。该算法将粒子群算法找到的当前最优值禁忌一段时间后再释放,以此避免算法陷入局部最优,即使算法暂时陷入局部最优,该算法跳出局优的能力也很强。实验表明,TPSO在收敛速度以及收敛精度方面都比基本粒子群算法有了很大程度的提高,特别对于多极值问题搜索效果非常好,可以很好的解决算法陷入局部最优的问题。
As the basic PSO may be easily trapped into local extreme,the taboo algorithm was brought into the basic particle swarm optimization and a new method called taboo particle swarm optimization（TPSO）was proposed.This algorithm releases the current optimal value after it was tabooed for several steps to avoid trapping into local extreme efficiently.Even when the algorithm was trapped into local extreme,it can jump out easily.The experiments results reveal that the TPSO is better than PSO in convergence velocity and convergence precision especially to multimodal functions.It is a very good solution to the local extreme problems.
出处
《陕西理工学院学报（自然科学版）》
2011年第1期85-90,共6页
Journal of Shananxi University of Technology:Natural Science Edition
关键词
粒子群算法
禁忌搜索算法
收敛速度
收敛精度
particle swarm optimization
taboo search algorithm
convergence velocity
convergence precision
