Abstract: The control strategy of the serial production system with machine breakdowns and stochastic demands is researched. First, minimization of the average total cost is taken as the optimization objective to establish an optimization model for serial production system’s control strategy based on（Q, r） inventory policy. And then considering the complexity for this problem to be solved, an improved Mesh Adaptive Direct Search（MADS-GA） is developed, in which the stagnation prevention strategy is adopted to avoid falling into local optimum and to improve the convergence rate. Finally, the proposed algorithm is adopted to solve the research case and the comparison with existing algorithms is also performed. Experimental results shows that, compared with the standard MADS（MADS-N） algorithm and the GA, the proposed algorithm is superior in the convergence rate and the optimization ability and can effectively optimize the control strategy of the serial production system, reducing the average total cost.

Key words: multi-echelon serial production system, machine failure, （Q, r） inventory policy, Mesh Adaptive Direct Search（MADS）, Genetic Algorithm（GA）, stagnation preventive strategy, simulation

摘要： 针对多级串行生产系统，考虑存在机器故障和随机需求，对生产及雇佣的控制策略进行研究。以最小平均总成本为目标，建立基于（Q，r）库存策略的优化模型；鉴于求解复杂性，提出一种改进的网格自适应直接搜索算法（MADS-GA），为避免陷入局部最优和提高收敛速度，算法中加入了停滞阻止策略；对研究实例进行求解，并与已有算法比较。结果表明，所提出的算法收敛速度和寻优能力更优，能有效优化串行生产系统的控制策略，降低平均总成本。

关键词: 多级串行生产系统, 机器故障, （Q, r）库存策略, 网格自适应直接搜索, 遗传算法, 停滞阻止策略, 仿真