Design&Implementation Of Production Data Analysis System In Auto Factory Stamping Based On Data Mining

Since our country implements the development strategy that informatization drives industrialization, enterprise informatization has got fast development accompany with the economic construction. Database management system has already been used widely in the field of industrial production, so as to make production information data which enterprise accumulates becomes more and more. Enterprise hopes to analyze these production information data in a higher level, so as to make full use of these data. Then some decision basis will be provided for aiding enterprise how to improve the management level and reduce production cost.This paper firstly introduces the concept, characteristics, classification and common technology of data mining, summarizes the generation time background of data mining and applications in various fields, and states the research status and development direction of data mining at home and abroad.Secondly, the steps of association rules mining and the details of classical algorithm and improved algorithm are analyzed. According to characteristics that If items are more and time complexity of algorithm is higher, the efficiency will be lower, the improvement of associate rule algorithm which suits for practical situation in stamping plant has been put forward combined with the case that stamping plant can explicit seccedent (refers to the property which react product quality condition) which is explored by association rules. Then the items of actual participation in calculation have been reduced, as a result that time and space consumption of algorithm has been reduced and system mining efficiency has been improved. Aim at the problems of production information management system in stamping plant that data is redundancy, irregular, inaccuracy and inconsistencies because of scattered data sources, the data warehouse catered to production data analysis theme is established.Which has provided high quality data for data mining, Simplified steps of data mining and reduced the time of data preprocessing before mining.At last, on the base the information management system in stamping plant and combined with SQL SERVER2005database, production data analysis system software is developed based on C/S structure by using delphi7. This analysis software screen is intuitive and understandability, simple to operate and is well expansibility.Which has provided decision basis for administrative staff to improve product quality and reduce scrap rate products.The main characteristics of this paper are that data warehouse catered to production data analysis theme is established and optimization of association rules algorithm is realized combined with the actual situation in stamping plant. Combined data warehouse and data mining technology, a data mining system which suits for analyzing production information data in stamping plant is designed.So as to come up with the strategy of improving product quality. The successful development and application of this system have done a beneficial attempt for analyzing automobile spare parts manufacturing data by using data mining and the problems that many varieties, more processes, mix production line and a large quantity of data have made data analysis difficulty.