Optimization of Enterprise Production Processes Based on Multi-Stage Decision-Making Models

This paper focuses on issues related to the inspection of spare parts, finished products, and the handling of defective items in enterprise production. Using methods such as sampling inspection and dynamic programming, a multi-stage mathematical model is developed to optimize production decisions. A statistical hypothesis testing scheme is designed for the sampling inspection of spare parts procurement, aiming to determine the optimal sample size. Dynamic programming models are introduced at each stage of the production process to explore the necessity of inspections and the impact of costs under six different scenarios, thereby formulating strategies. A multi-stage production decision model is constructed, treating production as a Markov decision process to solve for the optimal decisions. Considering fluctuations in the defect rate, confidence intervals are calculated using hypothesis testing, and Monte Carlo simulations are employed to generate the defect rate. The expected returns are then recalculated to increase enterprise profits.