Control Of Quality Process Based On Artificial Intelligence

With the evolution of market competition and consumption mode, the implication of quality has been expanded to the entire lifecycles of products, which produce much information needs to be integrated. On the other hand, automatic production processes require intelligent quality control. Driven by these demands, integration technology of quality information in the entire lifecycles of products and methodology of intelligent quality control are investigated in this thesis. The main contents are as follows:1. After analyzing quality information in products' lifecycles, the frameworks of quality information system based on XML technology are established. They provide efficient platforms for quality information integration.2. Through investigating the mechanism of quality control system based on MES (Manufacturing Execution System), the framework and function model of dynamic quality control system are constructed. Further, the online quality inspection system model based on SPC (Statistic Process Control) is advanced.3. The recognition methodology of quality process control patterns based on the BP neuron network is investigated in this thesis, which includes establishing the mathematical model for abnormal patterns of the quality process, constructing the diagnosis and analysis system model of the quality process, describing the algorithm of moving window recognition, designing, training and testing the BP network.4. This research further suggests a way to estimate the characteristic parameters of abnormal patterns of the quality process based on BP neuron networks. These parameters can provide further clues to reveal potential quality problems in manufacturing processes.5. After introducing rough sets theory in this thesis, a model of quality process deviation diagnosis based on rough sets theory is presented and the efficiency of this method is exemplified.