针对最小化生产节拍的第二类混流U型装配线平衡问题, 构建了混合整数规划模型, 并设计了改进型遗传算法求解. 在遗传算法的解码过程中, 采用3种搜索方式将工序分配到工作站, 比较了3种搜索方式下的工作站时间, 并参照期望生产节拍值筛选出最优的工作站分配方式, 再根据分配方式的寻优情况判断是否自动更新期望生产节拍值. 通过大量的标准算例, 证明了改进型遗传算法的有效性. 最后, 结合实际案例分析, 再次验证了混合整数规划模型和改进型遗传算法的有效性.

To solve the type-Ⅱ balancing problem of a mixed-model U-shaped assembly line to minimise the cycle time, a mathematical model was established and an improved genetic algorithm was designed. In the decoding process of the genetic algorithm, three search techniques were used to assign tasks to workstations, and workstation times under these search techniques were compared. The optimal assignment was selected according to the expected cycle time. The task assignment result determined whether to update the value of the expected cycle time automatically. The performance of the proposed improved genetic algorithm was proved by a set of benchmark instances. Finally, a practical balancing problem was analysed and effectively solved by the proposed method.