Study On Quality Control Forecasting Of The Purchasing Parts Based On Data Mining

Quality is of great importance to any enterprises. It is very important to control the purchasing parts'quality in the automobile production. If we can predict its quality problems, we can improve the automobile's quality a lot. This thesis mainly analyzes and studies Data Mining and Time Series Analysis according to the realistic situation of one Shanghai Automobile Factory.According to the problems exist in the purchasing parts, we introduce Data Mining and Time Series Analysis to analyze and predict the historic checking data of the purchasing parts. First, we introduce the concept of Data Mining and Time Series Analysis, use the study methods and arithmetic of Data Mining. We use those methods to deal with the historic data. We adopt the quality control data in actual purchasing parts measurement to establish the AR model, and then we establish a set of effective purchasing parts quality control system to have a Time Series Analysis based on the model we have constructed. Also JSP and SQL Server 2000 Database technologies are used to create the information flow management system. We use sample data identify model parameters and take the dynamic prediction on the parts quality parameter in future according to the identification model.What we discussed in the thesis comes from the practice of the factory and it has been proved to possess good application value in real practice. Now, the system provides useful quality reports to the Quality Control Department for them to analyze where the problem is.