关键词: 质量控制, 贝叶斯估计, 蒙特卡罗方法, 马尔可夫过程, 仿真

Abstract: To solve the problem that the data generated by autocorrelative process couldn’t meet the basic assumption of variables’ independence in quality control， Bayesian quality control method was introduced. And the posterior distributions of parameters in quality model were simulated via Markov chain Monte Carlo based on Gibbs sampling， by which the Bayesian statistical quality control model was set up with the autocorrelative data. Residual series in the model were independent so as to satisfy the necessary assumptions in quality control charts. The Bayesian model could avoid incorrect alarming when the process was under control， or not alarming when the process was out of control. Results showed that the Bayesian method was an effective tool to monitor the quality control of autocorrelation process.

Key words: quality control, Bayesian estimation, Monte Carlo method, Markov process, simulation