多品种小批量生产模式下共轭贝叶斯动态质量控制方法

Conjugate Bayesian Dynamic Quality Control Method under Multi-variety and Small-batch Production Mode

以多品种、小批量生产模式下数据量少及无法实现质量控制的问题为研究对象,采用共轭贝叶斯理论实现产品动态质量控制为研究目的。利用成组技术相关理论从相似工序的过去某段时间的生产批次数据中抽取数据,通过统计变换数据并再次经过Lilliefors检验、F检验确定为先验信息,然后利用时间序列指数加权理论对历史数据各批次赋予权重,通过贝叶斯正态共轭方法建模,计算工序能力指数实现工序质量衡量。仿真结果表明,随生产的不断推进,数据量的增多,模型输出的控制限不断靠近实际值,控制效果不断提高。说明所建模型能够实现产品工序质量的及时追踪和反馈,能有效解决多品种、小批量工序质量控制中出现的质量控制问题。

Taking the problem of less data unable to realize quality control in multi-variety and small batch production mode as the research object,the conjugate Bayesian theory is used to realize product dynamic quality control.The relevant theory of group technology is used to extract data from the production batch data of similar processes in a certain period of time in the past,and the data are statistically transformed and determined as prior information through Lilliefors test and F test;then each batch of historical data is weighted by time series index weighting theory,and the process quality is measured by Bayesian normal conjugate method.The simulation results show that with the continuous progress of production and the increase of quantity,the control limit of the model output is constantly close to the actual value,and the control efficiency is constantly improved.Describe the model established in this paper can achieve timely tracking and feedback of product process quality,and it can effectively solve the quality control problems in the process of quality control of multivariety and small batch.