| This paper is based on reading lots of related domestic and overseas articles, analyzing the mathematical former and the holistic control structure of copper rolling control system, especially dicussing and researching deeply on theory of mass flow control, and discussing some problem in the process of controlling. And researching on the theory and design procedures of BP neural network predictive control based on the theory of mass flow control. I give my own view and opinion for the BP neural network applying on the mass flow control system, and provide feasibility of the BP neural network applying on the mass flow control system.Aiming at character of changing parameter for related measuring parameter in the needed predicted signal, this paper provides the idea of the BP neural network predicted control based on the theory of the mass flow control. This idea's essence is to use the predicted signal by neural network of error feedback model, then using optimized arithmetic to educe control rule, consequently implementing predicted control on changing parameter system. The selected neural network is BP network that have favorable ability of function approaching. First identifying the controlled object outside of the line; after the model reaching definite precision, and then getting the predicted model on the line; finally you can get optimal control rule through function with minimization performance figure. And using this control rule in the model of control system. Analyzing the feasibility of applying this generalized predicted control on like copper rolling so fast operating system, and makes the field of application of predicted control more widely.The aim of this paper is to analyze the practical application feasibility of BP neural network based on the theory of mass flow control. In this paper, I put forward my own views and opinions, and establish definite basement for using BP neural network apply on practical system in the future.
|
PDF Full Text Request