基于多参数优化ACO的装备零件回收路径  

Equipment parts recovery path based on multi-parameter optimization of ACO

作　　者：孟小玲 温海骏[1] 祝锡晶[1] 曾艾婧 邵延君[1] MENG Xiao-ling;WEN Hai-jun;ZHU Xi-jing;ZENG Ai-jing;SHAO Yan-jun(School of Mechanical Engineering,North University of China,Taiyuan 030051,China)

机构地区：[1]中北大学机械工程学院,山西太原030051

出　　处：《计算机工程与设计》2021年第5期1278-1285,共8页Computer Engineering and Design

基　　金：国家自然科学基金面上基金项目(51975540);山西省研究生创新基金项目(2020SY349);山西省自然科学基金项目(201701D121079、201801D121185);2018年先进制造技术山西省重点实验室开放基金项目(XJZZ201808)。

摘　　要：为提高蚁群算法运行效率,缩短回收车辆路径长度,提出用响应曲面法优化算法参数的方法。选取算法中的4个优化变量和1个响应变量进行中心复合实验,运用Design-Expert软件对实验数据进行方差分析并建立回归预测模型,通过3D响应面图分析各参数及其交互作用对路径长度的影响,得到蚁群算法最优参数组合。实验结果表明,算法优化参数后能大幅提高寻径过程中有效蚂蚁数量,在寻优能力、收敛效果及稳定性方面都优于参数优化前,为进一步讨论装备零件回收再制造的车辆路径物流问题提供了一种优化思路。To improve the operation efficiency and shorten the length of vehicle path of ant colony algorithm,a method of optimizing algorithm parameters using response surface method was proposed.Four optimization variables and one response variable were selected for the central compound experiment.The software Design-Expert was used to analyze the variance of the experimental data and establish the regression prediction model.The influence of each parameter and its interaction on the path length were analyzed through the 3D response surface graph,and the optimal combination of parameters of the ACO was obtained.Experimental results show that the optimized algorithm greatly increases the number of effective ants in the search process,and the optimization ability,convergence effect and stability are improved.It provides an optimization idea for vehicle routing problem of equipment parts recovery.

关 键 词：废旧装备零件 回收路径 蚁群算法 多参数优化 响应面法 最优路径 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]