Adaptive Control Based On Artificial Neural Networks Study Of The Injection Molding

Requests on product quality become higher and higher with application fields of plastic injection molding products change. In molding process, process variables directly impact flow state of melt in cavity and final quality of part. Precondition of improving part quality is to get and keep process variables optimization.In this paper, we mainly carried out the research on process control and optimization of process variables in plastics injection molding process. Some modeling approaches and real-time control strategies based on ANN are studied. A real-time intelligent system for controlling the injection molding process parameters, which makes product properties remain the optimum state throughout is realized by constructing a self-adaptive dynamic model. The main contributions are summarized as follows:Firstly, orthogonal array experiment was done to optimize process variables. We got the relative importance of various factors and the optimal factor level combination and chose the most importance process variables as control variables in process control. And offers the self-adaptive dynamic model some dynamic rules about the adjustment sequence of process parameters by orthogonal method and some empirical rules about the adjustment orientation of process parameters by studying specialist knowledge.Secondly, as a new intelligent control method, artificial neural networks (ANN) with nonlinear mapping and high parallel information processing capabilities pave a new way to solve the problem of identification and control of nonlinear systems. This thesis, on the basis of the research on structure and learning algorithm of neural network, BP network was used in injection molding process control by constructing the object model in theory.Finally, model reference adaptive control was chose as process control method. The results show that the system reduces notably the ratio of disqualified products and is more practical.