Research On Quality Control Of Process Under Small Batch Manufacturing

At present,the application environment of the traditional statistical control theory is a continuous mass production,and when the control method is applied to the quality control of the small mass production process,it will easily lead to the emergence of the problem of control distortion in the production process.The main reasons are as follows: on the one hand,it is difficult to obtain a large number of quality characteristic data related to current manufacturing information in small batch production,so that the traditional method can not build a monitoring model for the process quality control.On the other hand,the false alarm probability of the controlled drafting is subject to the current sample under the condition of small batch production.Capacity has a great influence.Therefore,in view of the distortion of the quality control in the production process under the small batch environment,the quality control method of the small batch production process is proposed in this paper,and the effective control of the quality of the production process is realized.In view of the shortage of sample size in small batch production,the traditional maximum likelihood estimation method can not effectively monitor the quality of the production process.This paper presents a statistical process control model based on the conjugate Bayesian estimation.Based on the total mean value obeys the conjugate normal distribution and the total variance obeys the conjugate inverse Gamma distribution,the prior information is selected from the historical information by using the matter element similarity theory,the Lilliefors normality test,the Bartlett variance homogeneity test and the ANOVA variance analysis,and then the overall score is given with the current sample information and the conjugate Bayes theory.The calculation method of the process capability index and the control limit of the quality control chart in small batch production is proposed.This method can effectively reduce the distortion problem that only uses the current small sample size to estimate the unknown distribution parameters.At the same time,the method can automatically adjust the current sample with the change of the current sample size.The proportion of information and prior information in Bayesian parameter estimation.Finally,the effectiveness and practicability of the quality control model established in the production process is verified by Matlab simulation and case study.Aiming at the problem that the false alarm probability of the control graph is influenced by the current sample size and can monitor the quality fluctuation of the production process at the same time,this paper proposes a joint control chart based on the variable control limit of Shewhart-EWMA.The control performance of the traditional single and joint control graph is compared.Based on the Shewhart mean control chart of the previous small sample,the relationship between the false alarm probability and the sample number of the joint control chart is analyzed,and the EWMA and Shewhart standard can be maintained according to its relation,the t distribution and the F distribution.The calculation method of the control limit for the false alarm probability of the quasi differential control chart makes the joint control chart of the Shewhart-EWMA variable control limit more suitable for the quality monitoring of the small batch production process.