中文摘要：

混流装配线的物料配送是支撑装配线运作的复杂系统。配送准确、及时不仅能够保证生产不断线,还能大大提高生产效率,发挥混流装配的优势。但是配送环节中往往存在工位货物需求量,工位预约到货时间和车辆运输时间不确定等因素,使得传统路径规划模型不能真实地反映现场情况,反而因种种异常导致事倍功半。对不确定因素进行考虑,建立模糊信息条件下的机会约束规划模型,并改进传统混合智能算法去求解模型。算法设计上,采用轮盘赌启发式算法来缩小初始解的搜索范围,在交叉算子中提出用广义海明相似度概念来区分染色体的相似程度,在进化过程中采用双选择双变异流程来加快算法的收敛速度。通过实例证明该算法处理不确定因素的可行性和高效性,并对模糊参数中关键因子的置信度选择不同值进行对比分析,给出选择建议。

英文摘要：

Material distribution of mixed-model assembly is a complex system for supporting assembly line operation.Accurate distribution in time can ensure continuous production and greatly improve production efficiency.However,there often exist uncertain factors,such as work station demand for goods,appointed goods arrival time,and vehicle transportation time,so the traditional route programming model cannot truly reflect the field situation and on the contrary will cause low productivity.The uncertain factors are taken into consideration,the chance constraint programming model under the condition of fuzzy information is built,and traditional hybrid intelligent algorithm is improved to solve this model.In the selection of initial solution,roulette heuristic algorithm is adopted to reduce the search range.In the crossover operator,generalized hamming similarity degree is used to distinguish the similarity degree of two chromosomes in order to avoid inbreeding.Duplicate selected and mutated operator is chosen for increasing the convergence speed.A practical example proves the feasibility and high efficiency of the algorithm in dealing with uncertain factors,and different values of confidence of key factors in the fuzzy parameters are selected for comparison and analysis,thereby giving the suggestion on selection.