Ant-colony-optimization strategy for assembly sequence planning

A directed assembly graph is introduced to represent the precedence relationship between components. After that, an assembly relation matrix is established. Then, an Ant Colony Optimization (ACO) strategy is presented to generate assembly sequence. The ACO algorithm is improved and is suitable to the task; a selectable set is established according to precedence relationship; a penalty-evaporation mechanism is proposed to punish the infeasible assembly sequence, which helps the ants construct valid solutions; the way of updating pheromones is arranged specially to fit the characteristics of an assembly sequence plan. Finally, an example illustrates the effectiveness of the strategy.