Production Decision Optimization Based on Markov Decision Process (MDP) and Genetic Algorithm

This study aims to focus on multi-stage decision optimization in enterprise production through models such as Markov Decision Process (MDP) and Genetic Algorithm. First, for the sampling and testing of spare parts, this paper determines the minimum sample size to determine whether the rate of defective spare parts exceeds the nominal value by means of a hypothesis testing model and sample size calculation, so as to provide the enterprise with a reasonable sampling and testing plan. Secondly, a multi-stage profit decision-making model is constructed, considering each cost and other factors, and adopting the traversal method to optimize the production decision. Finally, considering the complexity of the production process, this paper evaluates the long-term impacts of different strategies based on MDP and combines genetic algorithms to solve the optimal production decision. Through comprehensive comparison, the optimal decision scheme is finally derived. By systematically evaluating and optimizing the decisions of each production link, this study effectively copes with the complex decision-making situations faced in production.