Research On Multi-Stage Production Inspection and Decision Optimization Based on Dynamic Programming

Abstract

This paper addresses the quality control and cost optimization issues in the multi-stage production of electronic products by constructing a multi-stage production strategy model and dynamic adjustment mechanism based on dynamic programming. By quantifying the cumulative effect of defective rates, it was calculated that the cumulative defective rate for semi-finished products reaches a maximum of 0.271, while that for finished products reaches 0.569. An optimal path was designed: "Inspect seven parts, not inspect all semi-finished products, and not inspect but disassemble finished products." This reduced total cost to 88.2392 yuan, with only 5.6% of the loss due to defective products. The dynamic adjustment mechanism employs Wilson interval improvement estimation, which reduces the error by 23.8% with a sample size of 10 and avoids losses of 32 yuan in the scenario of a sudden change in defective rate. Simulation data shows that compared with the traditional strategy, the total cost is reduced by 14%, the qualified finished product rate increases by 5.2 percentage points, and the proportion of inspection costs decreases by 7.5%, verifying the model's effectiveness in balancing quality and cost.