Research on Quality Control Strategies Based on Hypothesis Testing and Optimisation Models

Abstract

This study proposes a multi-level quality control strategy based on hypothesis testing and optimization models, aiming to balance quality inspection and cost control in modern manufacturing. Firstly, a sample size estimation model based on hypothesis testing is constructed and combined with a two-stage sampling optimization strategy. This approach leverages statistical inference and phased sampling decisions to significantly enhance the scientific validity and resource utilization of inspection decisions. Secondly, a revenue optimization model based on 0-1 programming is designed, innovatively incorporating the concept of disassembly cycles. A greedy algorithm is employed to dynamically adjust inspection schemes, approaching global optimality through locally optimal choices. Results indicate that avoiding disassembly cycles is the optimal strategy for maximizing revenue. Lastly, a cost control model based on dynamic programming is proposed. By decomposing costs into sub-items such as spare parts procurement, inspection, assembly, and disassembly, and utilizing the multi-stage decision-making characteristics of dynamic programming, the model achieves minimum cost strategies. The reliability and stability of these models are verified through Monte Carlo optimization methods and extensive data simulations. The results demonstrate the feasibility and applicability of this strategy in complex production environments, providing a scientific basis for decision-making in dynamic manufacturing contexts.