面向多人同时作业的拆卸序列规划  

Disassembly sequence planning for multi-people simultaneous operation

作　　者：蔡凯骏 张伟明[1] 张梅军[2] 季立[2] 赵鸿飞[3] CAI Kaijun ZHANG Meijun ZHANG Meijun Jl Li ZHAO Hongfei(Department of Petroleum Supply Engineering, Logistical Engineering University, Chongqing 401311, China College of Field Engineering, PLA University of Science and Technology, Nanjing 210007, China 73016 PLA Troops, Zhenjiang 212416, China)

机构地区：[1]后勤工程学院军事供油工程系 [2]解放军理工大学野战工程学院 [3]中国人民解放军73016部队

出　　处：《计算机集成制造系统》2016年第12期2767-2777,共11页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金资助项目(51175511)~~

摘　　要：为生成面向多人同时作业的并行拆卸序列、提高求解的效率与质量,在分析现有方法不足之处的基础上,提出基于改进蚁群算法的面向多人同时作业的拆卸序列规划方法(Disassembly Sequence Planning For Multipeople Simultaneous Operation,DSPMSO)。针对该方法的特点,基于零件分层图对于零件配合关系以及可拆卸性的表达,提出改进的蚂蚁搜索方式,并采用动态候选表避免无效序列的产生;定义了零件拆卸所需人数与拆卸等待时间,提出以考虑等待时间为主的多人拆卸成本模型;针对基本蚁群算法求解复杂装配体拆卸序列时算法求解效率与解的质量难以兼顾的问题,提出算法的分阶段迭代策略:根据路径信息量确定算法迭代阶段,使蚂蚁具有不同的选路策略,提出与之对应的蚂蚁信息素自适应更新机制,使算法在求解的效率与质量之间取得较好的平衡。通过实例对关键参数的取值进行讨论,并验证了算法各项优化措施的有效性。To generate parallel disassembly sequence for multi-people simultaneous operation and improve the solving efficiency and quality, a method of Disassembly Sequence Planning for Multi-people Simultaneous Operation （DSPMSO） based on improved ant colony algorithm was proposed. Aiming at the characteristics of DSPMSO, based on the parts＇ assembly relationship and dismountability expressed in the stratified graph, the improved ant colony searching method was proposed. By using the dynamical candidates＇ strategy, the generation of invalid sequence was avoided. The required people and waiting time to remove a part was defined, and a multi-people disassembly cost function which mainly considered the waiting time was developed. Aiming at the difficulty to balance the efficiency and quality of solution when solving the complex assembly＇s disassembly sequence based on traditional ant colony algorithm, a phased iteration strategy was proposed. The phase of iteration was determined by the route＇s pheromone, which made the ants have different routing strategies. The ant adaptive updating rules of pheromone related to phased iteration strategy was proposed to achieved a better balance between the efficiency and quality of the algorithm＇s solution. Through a case, the value of key parameters was discussed and the effectiveness was proved.

关 键 词：多人同时作业 拆卸序列规划 改进蚁群算法 分阶段迭代 自适应更新 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]