Multi-stage Decision-making and Optimisation Model Based on Dynamic Programming

Enterprises often involve multi-stage decision-making requirements in spare parts procurement, assembly, testing and after-sales service, this paper proposes a multi-stage decision-making and optimisation model based on dynamic programming, which aims to help make scientific and reasonable decisions. Firstly, a binomial distribution model is used to establish a sampling strategy to assess whether the quality of parts meets the purchasing criteria. Subsequently, a cost-effective dynamic programming model is constructed to explore how to make decisions at each stage of a complex production process with multiple parts and processes, under the conditions of defective rate transfer. Finally, a total cost function is constructed and the overall profit is calculated by combining it with a revenue model. Finally, we demonstrate the effectiveness of the sampling scheme and the cost-benefit optimal scheme through real case studies. Moreover, we compare with traditional global planning methods to show the advantages of dynamic planning in dealing with large-scale problems, coping with dynamic changes, and optimising resources. In the future, we need to give more consideration to optimising the accuracy of the input data, as well as taking into account the dynamics of the supply chain, in order to make decisions more scientifically sound.