Research on Multi stage Production Process Optimization Based on Bayesian Inference Model

Abstract

Supply chain optimization not only affects the production efficiency and product quality of enterprises, but also directly impacts their cost control and profitability. This article focuses on the spare parts provided by suppliers, the multi-stage production process of enterprises, and quality control during the production process. A Markov decision process was used to construct an optimization decision model aimed at improving production efficiency and reducing defect rates through sampling testing and cost optimization. This provides decision-making support and a basis for optimizing the production process for enterprises.This article makes a decision on whether a company should purchase a batch of spare parts with a defect rate not exceeding a certain nominal value while minimizing the number of inspections. Describing the sampling process with a binomial distribution, establishing a confidence level based sampling detection model to estimate the defect rate, using a one-sided inspection method to determine whether the defect rate of spare parts exceeds the nominal value by minimizing the number of inspections, and making the optimal decision for the enterprise. Based on this, the acceptable defect rate for enterprises is limited to 0.05~0.2. Using statistical methods, when the reliability is 95%, the defect rate is 0.199, corresponding to the minimum sample size that needs to be tested, which is 104; When the reliability is 90% and the defect rate is 0.063, the corresponding minimum sample size to be tested is 335. At the same time, the sequential probability ratio test was used in the article to optimize the model. By selecting a small number of samples and making a decision on whether to continue sampling based on the results, the number of experiments was saved under the same reliability.