关键词: 按订单装配(ATO), 优化控制, 马尔可夫决策过程(MDP)

Abstract: This paper considers an Assemble-To-Order(ATO)system with capacitated inventory that produces n components with a single assembled product. Based on Markov Decision Process(MDP)framework, we create optimization algorithm to study the optimal production and inventory policy of ATO system. The results show that the optimal policy is state dependent, and the strategy of one component depends on the inventory of others. That is, the optimal production policy can be described as a series of base-stock levels, and optimal allocation policy can be described as a series of rationing levels. Both these two types of levels are increasing with the on hand inventory of other components. The dynamic programming theory, optimal control theory and numerical calculation method are used to study the existence of the optimal control strategy, and optimal value calculation. The project will build the decision model that corresponds better to the practical ATO system, establish the theory and carry out the experimental validation. The influence of system parameters on optimal policy is also proposed in this work.

Key words: assemble-to-o-rder(ATO), optimal control, markov decision process(MDP)