生产过程中的决策问题

Decision-Making Problems in the Production Process

摘要: 为解决电子产品生产中零配件采购质控、全流程检测决策及不合格品处理优化问题，本文以某企业畅销产品为背景，围绕采购、装配、检测、不合格品处理环节，构建多维度数学模型。针对采购抽样检测，结合假设检验等设计精准且成本效益优的方案；针对全流程，用模拟退火算法寻最优解，验证“全环节检测 + 不合格成品拆解再利用”为最优组合；推广至复杂场景，构建次品率传递模型。实验显示，模型在5类情境降本10%~15%，方法可推广至多领域，具有实践价值。

Abstract: To address issues in electronic product production—including quality control of spare parts procurement, full-process inspection decision-making, and defective product handling optimization—this study, based on a best-selling product of an enterprise, constructs multi-dimensional mathematical models focusing on procurement, assembly, inspection, and defective product handling. For procurement sampling inspection, it designs an accurate, cost-effective plan using hypothesis testing; for the full process, it adopts the simulated annealing algorithm to find the optimal solution and verifies “full-link inspection + disassembly and reuse of defective finished products” as the optimal combination. Extended to complex scenarios, a defect rate transfer model is built. Experiments show the model reduces costs by 10%~15% in 5 scenarios, with methods applicable to multiple fields and strong practical value.