拆卸序列规划, 干涉矩阵, 支撑矩阵, 蚁群算法

In order to solve the combinatorial optimization problem of disassembly sequence planning, an ant colony algorithm-based approach was presented. A multi-objective mathematical model was established which was expressed by matrixes and contained geometric constraints, stability constraints and disassembly tool constraints. Interference matrix was used to analyze the feasible disassembly direction, support matrix was used to analyze the stability, and disassembly tool lists were used to analyze the tools used during operations. In order to find the optimal disassembly sequence more quickly, redirection factor, tool change factor and stability factor were added to the state-transition rule. Local and global pheromone were updated after operations so as to avoid a local optimal solution, to accumulate pheromone according to the quality of solutions and to guide ant colony in searching the approximate/optimal solutions quickly. Finally, the disassembly sequence of CRH3 bogie wheels was used as a case research to show the feasibility of the algorithm.