改进型混沌PSO算法及其在VRP中的研究

Improved Chaos Particle swarm Algorithm on VRP

收敛速度、精度和全局搜索能力对于任意改进的智能化算法都非常重要。针对现有粒子群搜索算法在后期易陷入早熟、局部最优和收敛速度变慢的问题,论文提出了一种改进粒子群局部搜索能力的优化算法。改进混沌PSO算法在粒子群算法的基础上,提出对每次迭代位置求取平均最优解用来代替个体最优解并对求解后的最优位置进行混沌优化,同时加入收缩因子提高收敛速度,用来保证算法的平衡性和全局收敛性。通过对车辆路径问题的仿真实验结果表明,该改进算法在寻优精度和全局收敛能力方面优于参考文献中其他算法,其对于解决车辆路径问题是一种有效方法。

Convergence speed,accuracy and the ability of global search is very important for intelligent algorithms that can be improved by any way.Due to the existing PSO search algorithm is liable to cause some problems in the later,such as the precocity,local optimum and slow convergence speed.An improved algorithm is proposed in this paper,it can improve the local search ability of particle swarm.On the basis of particle swarm optimization（PSO）algorithm,chaotic PSO algorithm is improved.Specifically,it presents to get the average optimal solution for the position of each iteration instead of individual optimal solution and to do chaos optimization for the optimal position after processing.At the same time,it adds the shrinkage factor to improve the convergence speed to ensure balance and global convergence of the algorithm.The simulation results about vehicle routing problems show that the improved algorithm is better than that of refs,especially in optimization accuracy and global convergence ability.And it is also an effective method to solve the VRP problem.