一种对粒子群算法惯性权重的改进

针对粒子群算法搜索精度不高,特别是在处理高维复杂问题时极易陷入局部最优的不足,文章提出一种动态扩散并结合交叉因子的改进粒子群优化算法（DMPSO）,对惯性权重进行调整,对其取值范围做了进一步的研究,在必要的时候对整个种群的粒子进行重新扩散,并应用于粒子群算法的改进。实验结果表明,新算法的全局搜索能力、收敛速度、精度及稳定性均有了显著提高,而且能更有效地进行全局搜索。

An Improvement for Inertia Weight of Particle Swarm Optimization

For the poor search accuracy of PSO,particularly the deficiency of easy to trap in local optimum in dealing with complex high-dimensional problems,this paper presents an improved particle swarm optimization （DMPSO） by dynamic diffusion and combined with cross factor.The inertia weight is adjusted and its value range is further researched,the particles of the whole population diffuse again when necessary,which is applied to improve particle swarm optimization.The experiment results show that the new algorithm has significant improvement in global search ability,convergence rate,accuracy and stability,and can conducts global search more effectively.