基于遗传算法特性的混合粒子群算法求解TSP问题  

Hybrid particle swarm optimization algorithm based on genetic algorithm characteristics for solving TSP problem

作　　者：陈琳[1] CHEN Lin(The Internet of Things and Artificial Intelligence College,Fujian Polytechnic of Information Technology,Fuzhou 350001,China)

机构地区：[1]福建信息职业技术学院物联网与人工智能学院,福建福州350001

出　　处：《白城师范学院学报》2024年第5期73-78,共6页Journal of Baicheng Normal University

摘　　要：为解决粒子群算法在旅行商问题上的收敛速度慢和路径最优化选择的问题,提出了一种新型的基于遗传算法特性的混合粒子群算法,对旅行商问题的最优路径进行规划.根据种群比例原则与迭代前的路径进行交叉、变异、复制等操作,建立了具有遗传算法特性的混合粒子群算法,并用于求解burma14问题.结果表明:相比传统的粒子群算法和模拟退火-禁忌搜索算法,混合粒子群算法在求解burma14问题中收敛时间与最优路径等指标上都有明显的优势,且随着迭代次数与种群个数的增大,算法的最优解逐渐减小;当最佳参数为种群个数150,迭代次数300时,最优解为30.179 424.To solve the problems of slow convergence speed and path optimization selection of particle swarm optimization algorithm in traveling salesman problem,a novel hybrid particle swarm optimization algorithm based on genetic algorithm characteristics is proposed to solve the optimal path of the traveling salesman problem planning.Based on the principle of population proportion and the path before iteration,perform crossover,mutation,replication and other operations to establish a genetic algorithm,a hybrid particle swarm algorithm with genetic algorithm characteristics was used to solve the Burma14 problem.The results showed that compared to traditional particle swarm algorithms,convergence time and optimal path of hybrid particle swarm optimization algorithm in solving the Burma14 problem,along with simulated annealing taboo search algorithm,there are obvious advantages in indicators,and as the number of iterations and population increases,the optimal solution of the algorithm gradually decreases.When the optimal parameter is a population size of 150 and the number of iterations is 300,the optimal solution is 30.179424.

关 键 词：混合粒子群算法 TSP问题 路径规划 影响因素 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]