A max-min ant colony system for assembly sequence planning

An improved ant colony optimization (ACO)-based assembly sequence planning (ASP) method for complex products that combines the advantages of ant colony system (ACS) and max-min ant system (MMAS) and integrates some optimization measures is proposed. The optimization criteria, assembly information models, and components number in case study that reported in the literatures of ACO-based ASP during the past 10 years are reviewed and compared. To reduce tedious manual input of parameters and identify the best sequence easily, the optimization criteria such as directionality, parallelism, continuity, stability, and auxiliary stroke are automatically quantified and integrated into the multi-objective heuristic and fitness functions. On the precondition of geometric feasibility based on interference matrix, several strategies of ACS and MMAS are combined in a max-min ant colony system (MMACS) to improve the convergence speed and sequence quality. Several optimization measures are integrated into the system, among which the performance appraisal method transfers the computing resource from the worst ant to the better one, and the group method makes up the deficiency of solely depending on heuristic searching for all parallel parts in each group. An assembly planning system AutoAssem is developed based on Siemens NX, and the effectiveness of each optimization measure is testified through case study. Compared with the methods of priority rules screening, genetic algorithm, and particle swarm optimization, MMACS is verified to have superiority in efficiency and sequence performance.