Research On Bayesian Control Charts For Small-Batch Production Mode

With the continuous development of the commodity market,consumer demand tends to be individualized and diversified,and the customer market for modern manufacturing enterprises shows the characteristics of rapid and unpredictable changes in demand.The multi-variety and small-batch production model is becoming more and more popular among domestic and foreign manufacturing enterprises.When the production process of products is affected by random or abnormal factors,the traditional statistical quality control methods are no longer applicable due to the small amount of data reflecting the characteristics of product quality and the lack of a priori information of samples in the small-batch production mode.Therefore,it is very important to choose a quality control technique suitable for small-batch production mode to monitor the production process effectively.This thesis investigates Bayesian quality control charts for the small-batch production model based on statistical process monitoring techniques and Bayesian theory.The main contents are as follows:(1)According to the characteristics of the small-batch production process,under the reasonable a priori information assumption,a reasonable prior distribution is selected for the quality control charts design according to the actual research background.The empirical analysis is conducted with a factory welding robot’s steel plate production line.The economic statistical design model of the variable sampling interval control charts is established according to the unit time loss function and statistical parameter restrictions.(2)For the Bayesian hierarchical model,the effects of covariance and random error terms are considered in the absence of displaying analytic expressions.The Bayesian hierarchical model parameter estimation results are obtained through multiple simulated sampling by the MCMC method.The main factors affecting the number of product defects are discussed.(3)The quantitative calculation method of the performance of the Bayesian defect number control chart is given.Then,the Monte Carlo simulation method is used to calculate the average operation chain length of the controlled and out of control state of the control chart.Combined with practical cases,the effectiveness and practicability of Bayesian control chart in small batch production mode are verified.