The Applying Research For Auto Production Quality Control Management

With the development of automobile manufacturing industry rapidly, product manufacturing process is an important link in quality control management, is also one of the main sources of product quality problem. So this paper adopts a more effective quality control system, to further improve the level of automobile manufacturing’s quality control management. Use modern methods of data collection and the thought of manufacturing execution system MES, combined with the quality AUDIT management, there is big improvement in the effectiveness of the quality data. That lay a solid foundation for the quality control management. Through a vast, timely and accurate quality test data to ensure the decision makers can fully grasp the quality status of the enterprise. At the same time, in order to the analysis of quality testing data, this paper studies the current AP clustering algorithms and the minimalist rule extraction algorithm, and respectively based on spectrum analysis and concept lattice to propose the two following optimization algorithms.In order to traditional quality control management have more complex factors, such as quality test data size, more attributes, this paper puts forward a kind of method Affinity Propagation clustering based on spectrum analysis(Affinity Propagation-based on spectral analyze, AP-SA). First, by using spectrum analysis technology will be distributed in high-dimensional nonlinear mapping data collection to the almost linear subspace, high dimensional data to low dimensional mapping process. At last, through the AP clustering algorithm for mapping in low dimensional space to cluster the data, and improve the AP algorithm in high dimensional space clustering performance. The simulation results show that this method is compared with the traditional AP algorithm, no obvious advantage in the low dimensional data, but with the increase of the experiment data’s sample size and dimension, this method in the high dimensional data reduces the clustering of time and at the same time, also ensures good clustering effect.In order to traditional statistical process of quality control management are complicated, the production quality decision rules is big and complex extraction and so on, this paper is put forward the minimalist rules extraction optimization algorithm based on concept lattice. The optimization algorithm use the relationship between Extended Indistinguishable Matrix and concept lattice to construct concept lattice model, make quality monitoring integrated with production closely. In quality decision rules mining, The optimization algorithm by the partial order relation between the concept nodes, direct to determine whether all the concept nodes from decision attribute set have the parent nodes, and then according to the parent node’s connotation to get the minimalist decision rules set. Not only give the enterprise to provide a visual and easily understandable minimalist rules set, improve the level of product quality control management, but also simplify the steps of the minimalist rules extraction. Simulation experiment results show that the optimization algorithm has certain stability at the same time, also improve the efficiency of extraction.Finally, the application of engineering instance verified this topic design is feasibility and effectiveness.