An Optimized Sampling Framework for Industrial Quality Control Using Bootstrap, Hypothesis Testing, And Monte Carlo Simulation

Abstract

With the increasing demand for quality control of spare parts in the context of intelligent manufacturing, minimizing the number of samples while ensuring high accuracy has become a core issue in industrial production. Traditional sampling methods typically rely on large sample sizes, leading to increased inspection costs, extended production cycles, and difficulty in balancing accuracy and efficiency. To address this challenge, this paper proposes a adynamic sampling framework that integrates Bootstrap resampling, hypothesis testing, and Monte Carlo simulation. By generating representative Bootstrap samples, conducting Z-test-based hypothesis testing to estimate defect rates, and optimizing sampling rates through Monte Carlo simulation-driven stochastic programming, the framework achieves minimal sample size while maintaining high detection accuracy. Validation through simulation experiments and empirical analysis using industrial datasets demonstrates that the proposed method reduces quality control costs by 30%-40% compared to traditional methods while preserving detection precision above 95%. This study provides an efficient and practical solution for quality control in resource-constrained manufacturing environments, with significant implications for smart production systems.