Research On The SPC Control Chart Under The Multi-specification And Low Volume Production Condition

With the development of market economy, in order to win in the market competition, the enterprise must do process control strictly. Statistical Process Control(SPC) as one of effective method has performed more and more function in the aspect of such as process stability judgement, abonormal finding, non-conforming product reducing.The study are processed in the key technologies of manufacturing process the recognition of abnormal patterns of the SPC control chart and the quality control method of multi-type & small-batch products. The function structure and system structure of the manufacturing-oriented quality control system are presented based on it.Firstly, the mathematical model for abnormal patterns of SPC chart is established in this thesis, and then the network model made of BP network is designed to recognize abnormal patterns of the quality control chart. Structural parameters of the network model are selected and tested.Secondly, the Bayesian forecasting theory is applied to the quality control of multi-type and small-batch products. The dynamic model, parameters estimation, intervention and application steps for the method are studied.Finally, combining with project of a valve manufacturer, manufacturing oriented quality control prototype system which describes the function flow, information model are given and then make it true.This thesis provides original quality control system and makes it true based on study of SPC chart combining with valve manufacturer'ERP project. It can not only recognize effectively abnormal patterns of the SPC chart but also make further judgement by using these patterns recognized by adding on SPC chart for multi-type & small-batch production which the original system missed. Most important, it can make a further control based on these data. Meanwhile, it can be used to monitor the quality fluctuation during the process of the multi-type & small-batch production quickly and efficiently combining with the objective data, subjective information and the model itself.The study of thesis solved the problem which existed in the process of multi-specification & small-batch quality control. So it has a good practical meaning to make the middle-small-type enterprise acclimatize itself to the market's development freelier than before.