基于多目标殖民竞争算法的随机型双边装配线

Balancing stochastic two-sided assembly line with multi-objective colonial competitive algorithm

针对随机型双边装配线平衡问题所特有的操作方位约束、位置约束、区域约束和同步约束,以最大化线效率、最小化平滑指数和最小化单位产品总成本为目标,构建了考虑多约束、多目标的数学模型。提出了一种新型的多目标混合殖民竞争算法求解该模型,设计了相应的帝国初始化、帝国内的同化、殖民竞争等操作,并将殖民竞争算法的全局搜索能力与延迟接受爬山算法的局部搜索能力有机结合,以更快获得更优的Pareto解。通过具体实例测试,并将结果与当前文献和快速非支配排序遗传算法进行比较,验证了所提算法的可行性和有效性。

In view of the special additional constraints such as operational direction constraints, positional con- straints, zoning constraints and synchronous constraints in the widespread stochastic two-sided assembly line balan- cing problem, the multi-objective and multi-constraints mathematical model aiming at the Line Efficiency （LE）, the minimization of Smoothness Index （SI） and the minimization of Total relevant costs per product unit （Tcost） was built. A novel Multi-Objective Hybrid Colonial Competitive Algorithm （MOHCCA） which combined the global search ability of Colonial Competitive Algorithm （CCA） with the local search ability of Late Acceptance Hill-Climb- ing （LAHC） algorithm was proposed to get better Pareto-solutions and the corresponding procedure of empire ini- tialization, empire assimilation and colonial competitive was designed for solving the balancing problem. Cases re- sults on the benchmark problems compared with the ones presented in the current literature and fast elitism Non- dominated Sorting Genetic Algorithm （NSGA-II） validated the effectiveness of the proposed algorithm.