基于最大-最小蚁群系统的装配序列规划

提出一种结合了蚁群系统与最大-最小蚂蚁系统优点的装配序列规划(Assembly sequence planning, ASP)方法。对近十年基于蚁群优化的ASP文献中采用的优化指标、装配信息模型、实例零件数等进行综述和比较。为提高序列的装配效率区分度，研究方向性、并行性、连续性、稳定性和辅助行程等5项指标的自动量化方法，将其融入到蚁群优化多目标启发式函数和适应值函数中。为提高对最优序列的搜索能力，以装配几何可行性为基础，从蚂蚁数量的确定、最大-最小信息素的界定、初始零件分配位置的绩效考核机制以及对并行零件组强制优化机制等方面，设计针对性解决方案，提出基于最大-最小蚁群系统的ASP算法。开发基于Siemens NX的装配规划系统AutoAssem。以阀门为实例，验证了算法内部各项优化措施的有效性，同时与优先规则筛选法、遗传算法及粒子群算法进行比较，分析该算法在运行效率和序列性能方面的优势。

Assembly Sequence Planning Based on Max-min Ant Colony System

An assembly sequence planning (ASP) method that combined the advantages of ant colony system (ACS) and max-min ant system (MMAS) is proposed. Several characteristics that adopted in the literatures of the ASP based on ant colony optimization (ACO) in the last decade are reviewed and compared, such as the optimization criterions, the assembly information models, the numbers of components in cases study. To identify good sequences more obviously, five optimization criterions are automatically quantified, including directionality, parallelism, continuity, stability and auxiliary stroke, and integrated into the multi-objective heuristic and fitness functions of ACO. To improve the search capability for the global best sequence based on geometric assembly feasibility, several measures are presented from the aspects of determining ant number, defining max-min pheromone, and the mechanisms of performance appraisal for initial components allocation and parallel components group enforcement. Then the ASP algorithm based on max-min ant colony system (MMACS) is proposed. An assembly planning system “AutoAssem” is developed based on Siemens NX platform, and the actual effectiveness of each optimization measure is testified through case study of a valve. Compared with priority rules screening, genetic algorithm and particle swarm optimization, the superiority of the algorithm in executive efficiency and sequence performance are analyzed.