基于混沌粒子群优化的新型VRP求解算法

标准粒子群优化算法(PSO)容易陷入局部最优,且精度较低、收敛速度慢,难以满足求解VRP的需求。本文提出了一种适用于求解VRP模型的新型混沌粒子群优化算法(CPSO)。该算法引入混沌序列,利用混沌对粒子的初始位置进行初始化,提高了样本的质量,并且对当前粒子附加混沌扰动,促使其跳出局部最优,提高了全局搜索能力,有利于在全局范围内寻找到最优值。实验结果表明,本文算法的收敛速度、精度及稳定性高于PSO算法,是一种有效的VRP求解算法.

Novel Chaos Based Particle Swarm Optimization Algorithm for the VRP Solution

Since the particle swarm optimization algorithm( PSO) is easily trapping in local optimum,has low precision and slow convergence rate, thus it is difficult to satisfy the requirement of the VRP solution.To solve these problems,a novel algorithm based on chaos particle swarm (CPSO) is proposed in this paper.This algorithm introduces chaos sequence and initializes the initial position of particles by using chaos. Accordingly,the sample quality is improved,and the chaos perturbation of the current particles conduces to avoiding local optimization,as a result of which the global searching ability is improved and it is advantageous for seeking the optimal value within the global scope.The experimental results indicate that the algorithm in this paper has better convergence speed,precision and reliability than PSO and it is an effective algorithm for VRP solutions.