基于多目标蚁群算法的复杂设备拆卸序列规划

针对复杂设备转化成混合图信息模型后面临的组合爆炸和搜索爆炸问题,提出了一种改进的Louvain算法,对其模型进行社团化处理。将拆卸工具、拆卸方向改变次数以及拆卸时间作为优化目标,构造了基于多目标蚁群算法搜索社团划分可行解;并得出由多个Pareto解构成的平衡解集。最后结合直升机旋翼,利用所提方法进行算法验证;对比未被社团化处理的多目标蚁群算法而得出的拆卸序列,在算法运行效率上提升了17.2%;且复杂设备中紧密度越小零件占比越多时,运行效率越高;对比单目标蚁群算法社团化后得出的拆卸序列,在迭代次数相同的条件下,更易得到较高效率的解。实验结果表明,在实际拆卸序列规划问题上,该方法为方案设定人员提供了更合理高效的选择。

Disassembly Sequence Planning of Complex Equipment Based on Multi-objective Ant Colony Algorithm

Aiming at the issue of the combinatorial explosion and search explosion problem after the complex equipment was transformed into mixed graph information mode, an improved Louvain algorithm based on community way was proposed. The disassembly tools, number of demolition direction change as well as the disassembly time has been regarded as the optimization target, and Multi-objective ant colony algorithm search feasible solution has been established and obtained several Pareto optimal solution of the set. Finally, combined with examples of Helicopter Rotor, compared with the demolition sequence which was obtained without matched processing, the efficiency of the algorithm has been enhanced by 17.2% under the proposed method. And a higher proportion of the smaller parts in the complex equipment, the higher the operating efficiency. A high efficient solution can be obtained more easily at the same conditions of the iteration compared to the demolition sequence which is obtained by community Single-Objective ant colony algorithm. The theoretical analysis and simulation results indicated that the proposed method provides a more efficient and reasonable choice for the programmer In the issue of actual planning demolition sequence.