Abstract: In order to improve utilization rates of vehicles， reduce makespan and total costs， an integrated optimization model of intercell scheduling and the vehicle routing was established to minimize makespan and total costs. Interrelationships between processing sequences of parts and transportation routes of exceptional parts were considered. A five-segment coding method including operation， machine， cell， vehicle and vehicle route was proposed. A MOEA/D was designed. To ensure feasible solutions in iterative processes of the algorithm， a binary tree based adjustment method and a vehicle capacity constraint adjustment method were proposed. The proposed model was verified through a case study. Results were compared to those of a common transportation strategy and a vehicle sharing strategy. The proposed integrated optimization model may yield better results on both makespan and total costs. 

Key words: cellular manufacturing, intercell scheduling, vehicle routing problem, multi-objective evolutionary algorithm based on decomposition（MOEA/D）

摘要： 为了提高跨单元调度中车辆利用率，降低运输成本和缩短完工时间，综合考虑调度方案中零件的加工顺序和异常件运输路径之间的相互影响，建立了一个以最小化最大完成时间和总成本为优化目标的跨单元调度及其车辆路径集成优化模型。提出了一种包含工序编码、机器编码、单元编码、车辆编码和车辆路径编码的五段式编码方法；设计了基于分解的多目标进化算法，针对算法迭代过程中可能产生不可行解的问题，提出了一种基于二叉树的车辆路径调整方法以及车辆载荷调整方法。通过实例分析，并将集成优化结果与常见运输策略结果、车辆共享策略结果进行了对比，集成优化模型在缩短最小完工时间和降低总成本方面可获得更优的结果，从而验证了所提集成优化模型的有效性。

关键词: 单元制造, 跨单元调度, 车辆路径问题, 基于分解的多目标进化算法