Abstract: An improved-ant-colony-algorithm strategy is presented for generating optimal assembly sequences that are feasible for assembling tools’ operation. The strategy makes use of a directed-graph based assembly model that records the precedence relations between the components, and represents fasteners and normal parts differently. To examine the feasibility of the sequences for tool’s operation, a quick checking method is developed based on potential interference information, which is established in an entire assembly model by simulating the tool’s movement and checking the interference between the tool and adjacent parts. Then, the ant colony algorithm is improved to generate optimal sequences with the guidance of precedence relations and the check of tool operation feasibility. Additionally, a penalty-evaporation mechanism is proposed to punish unfeasible solutions in order to help the ants constructing valid sequences. Finally, an example illustrates the effectiveness of the strategy.

Key words: Directed graph, Ant colony optimization, Assembly sequence planning, Assembly tool