改进蚁群算法在VRP问题中的应用及颜色Petri网实现  

Application of Improved Ant Colony Algorithm in VRP Problem and Its Colored Petri Net Realization

作　　者：郑文艳 赵丽敏 ZHENG Wen-Yan, ZHAO Li-Min(School of Information Management, Dezhou University, Dezhou 253023, China)

机构地区：[1]德州学院信息管理学院,德州253023

出　　处：《计算机系统应用》2018年第11期186-191,共6页Computer Systems & Applications

摘　　要：蚁群算法在解决车辆路径问题时存在运行速度慢等问题,基于此本文提出了一种自适应蚁群算法.该算法把客户需求等因素加入禁忌表,实时记录当前最优解,据此智能调整信息素的更新规则,同时调整了概率转移公式和可行解的构造方法,并建立了相应的颜色Petri网模型.最后利用VRP问题库中的几个经典实例与GA及其他改进蚁群算法进行了对比试验,验证了该算法既可以加快收敛速度,又可以避免局部最优,同时保证了最优结果的多样性.This research has proposed an adaptive ant colony algorithm and its colored Petri net approach, aiming to solve the slow operation problems existed in vehicle routing problems. The new developed method has improved the updating rule of pheromone, changed the formula of transfer probability and the constructed technique of feasible solutions, with the Petri net model conducted accordingly. The proposed method is finally verified compared to GA and some other ACOs through standard test cases, with the results demonstrated that the method developed effectively improves the convergence efficiency, refrains the searching from being trapped in local optima, and ensures the diversity of the final solutions.

关 键 词：蚁群算法 车辆路径问题 颜色PETRI网 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]