Design Of Supervision System And Research On Monitoring Methods For Injection Molding Process

Injection molding machine plays an important role in plastic industry, as the main equipment and typical representation of injection molding mechanism. Design of supervision system and research on monitoring methods for injection molding process will greatly improve the plastic production quality and promote the development of plastic industry.Based on the injection molding process, according to the research of lots of reference literature, it firstly introduces the injection molding craft and summarizes control system development and current process monitoring state. Then, corresponding to the injection process monitoring requirements, it builds the host computer monitoring system using QNX, one kind of real time operation system, and introduces the function of monitoring system, design and development. The practical run shows that the monitoring system preferably satisfies the process operation and monitoring intention, which provides rich data information for the theory research on process modeling and system analysis.Injection molding process, as one traditional multivariate and multi-stage batch process, is also the important application background of process monitoring, which has the increasing theory productions in monitoring. However, due to the process high dimensionality, complexity, batch-to-batch variation, its application in practice is not so satisfying, and precise injection control cann't be solved preferably. On the basis of comprehending multivariate statistical analysis knowledge, the paper further studies PCA and relative extended modeling methods and discusses the correlation changing with operation time focused on the multi-stage performances. It realizes the proper stage partition by improved clustering algorithm and then proposes a soft transition multiple PCA modeling and monitoring method based on batch process stages, which is demonstrated effectively with the application of injection molding process experiment and indicates the potentials for practical production.