The Monitoring And Fault Diagnosis Platform Of Injection Molding Process Based On Multivariate Statistical

Molding is a widely used method in plastic processing. Injection molding machine is the main processing equipment. Injection molding process is a typical nonlinear, dynamic and multi-stage batch process. The mechanism of Injection molding process which is complicated, and the complicit oi its operation is outweighed the continuous process. Therefore, many efforts have been devoted in injection process monitoring, fault diagnosis and the forecast method of quality in the industry have become the focus.This research has designed and investigated a web-based Injection molding process monitoring and fault diagnosis system software regarding to the injection molding process. The software includes four main functions: data flow, data storage, data operations and data display. The software has user-based data management, network distribution, uniform data interface and other new features by using Socket programming, multithreading, Ado.net, GDI+, Matlab&.Net and other key technology. The experimental platform makes the software suitable for practical application of advanced theoretical methods on the production of injection molding process. Meanwhile, it also helps junior researchers learn classical theory.In addition, with growing application of injection molding products in various fields, higher quality requirements of the products are required. However, in practical engineering applications, the quality control of injection molding machine has not been achieved. The main reason is the quality of closed-loop control cannot be measured sometimes, or.it has a serious time lag. This has become a bottleneck of quality control in injection molding industry.Therefore, in view of the date characteristics of the injection molding production process, this thesis investigated the application of the kernel function based on partial least squares method (KPLS) in injection molding process quality prediction. Based on PLS and MPLS quality prediction theories proposed foi batch process forecasts, a MKPLS method and analysis conclusions are provided here. Although the method has certain flaws, it concluded that the nonlinear issue is the breakthrough in monitoring, fault diagnosis and quality prediction method of the batch process. These conclusions direct the future of the theoretical methods.