关键词: Elman网络, 网络化制造, 相似制造, 专家系统

Abstract: Aiming at complicated and variable environment of networked manufacturing, how to settle the problems of monitoring dynamic and variable quality fluctuation, diagnosing the abnormal variation and adjusting the process at the right moment, is a difficult problem that faces the networked manufacturing enterprise in process quality control. An intelligent process quality control mode oriented to networked manufacturing, which integrates quality prevention, analysis, diagnosis and adjustment, and corresponding functional modules and framework, is put forward. This mode mainly deals with constructing and running dynamic quality control system, integrates theories and technologies such as the theory of similarity manufacturing, statistical process control (SPC), neural network. Furthermore, some of the key enabling technologies are studied in detail, including process quality analysis on-line based on similarity process, process quality diagnosis based on Elman and expert system of process quality adjustment. Finally, the prototype system of intelligent process quality control is developed by applying the development mode of JSP+Servlet and the network function interface of Java and Matlab, which is introduced as an example of some Heavy Machinery Ltd to demonstrate the rationality and validity of the method. It lays the foundation for realizing networking, intelligent and automatic process quality control.

Key words: Elman network, Expert system, Networked manufacturing, Similarity manufacturing