第15卷第2期 2017年4月
中国工程机械学报
CHINESE JOURNAL OF CONSTRUCTION MACHINERY
Vol. 15 No.2 Apr.2017
基于改进粒子群算法的自车身焊接路径优化
乐英，岳艳波
（华北电力大学能源动力与机械工程学院，河北保定071000）
摘要：为优化白车身焊接路径，提高焊接效率，提出一种改进粒子群算法，在传统粒子群算法思想的基础上，将算法导优过程分为追随和盘旋两部分，基于较近原则生成初始粒子，以减少种群规模，加快收速度；在追随部分，通过个体极值追随全局极值和随机原始参考值以贪重组的方式重新生成粒子，在增强算法局部寻优能力的同时加快算法的收做速度；在盘旋部分，采用多次局部调序的策略，通过随机调整粒子局部排列序，保证算法种群的多样性，防止陷入局部最优解；从种群进化代数和种群个体适应度函数值实现算法各参数的自适应调节，加快收敛速度；对粒子个体采取精英保留策略，保留最优粒子.算法通过Matlab平台实现，实验仿真结果表明，提出的改进粒子群算法对于中小规模的白车身焊点旅行推销员问题（TravellingSalesmanProblem，TSP)有良好的寻优能力，
关键词：焊接路径；改进粒子群；贪婪重组；多次局部调序；自适应调节
文章编号：1672-5581(2017)02-009908
中图分类号：TP183
文献标志码：A
ResearchonprocessplanningofweldingpathinBIwWbasedon
modifiedparticleswarmalgorithm
YUEYing，YUEYanbo
(School of Energy, Power and Mechanics Engineering, North China Electric Power University, Baoding 071000, Hebei, China)
Abstract; A modified particle swarm algorithm is proposed to optimize the welding path and improve the welding efficiency. Based on the traditional particle swarm optimization algorithm,the optimization process of algorithm is subdivided under the following and gyrating. In order to reduce the population size and improve the convergence speed,the proximity principle is introduced into particle species initialization;in the following part,in order to keep the algorithm good convergence and improve the convergence speed, the particle is re-generated by means of the greedy recombination of the individual extreme value following global extreme value and random original reference value;in the gyrating part,in order to keep the population diversity and prevent the local minimum, a multi-bit adjustment order strategy is introduced,the partial arrangement of particles is randomly adjusted; the adaptive update of algorithm parameters is realized according to the evolution stages and the fitness value of particle individuals,and the convergence speed is improved; the optimal particle individual is propagated using the strategy of keeping the best individuals. The modified algorithm is realized using Matlab, the simulation results show that the proposed modified particle swarm optimization has a powerful search capability in solving travelling salesman problem(TSP).
Key words:welding path;modified particle swarm optimization; greedy restructuring; multi-bit adjustment order;adaptive adjustment
基金项目：河北省自然科学基金资助项目(E2014502042)：中央高校基本科研业务费专项资金资助项目(11QJ61）
作者简介；乐英（1971—），女，副教授，博士.E-mail：yueying71@163.com 万方数据