Study On Application Of Predictive Control In Industrial Furnace Temperature Control

This paper studies application of predictive control algorithm in industrial furnace temperature control from the instance of glass furnace. In this paper, after introducing the research and development of intellective control algorithm of industrial furnace temperature control we simply analyze the basic theory and prepotency of predictive control algorithm. We create predictive model by analyzing parameters of glass furnace, and discuss a kind of predictive control algorithm proper for industrial furnace temperature control, which has rapid responsive-speed and good robust capability?Dynamic Matrix Control based on none-parameter model(for short DMC).Industrial furnace is applied widely in the Building Material Industry, and the quality of production depends on the temperature control technology. For Industrial furnace, especially for glass furnace, its controlled variable is lagging and its input variable is vulnerable to disturbing. Generally, glass furnaces have some special properties梟one-linear, circumstance changing with the passage of time and no-certainty, and are often multiple variable systems, thus it is very difficult to build actual numerical model. Even if we can build the model for some objects, it must be very complicated and be difficult to bring about good control effect. Self-adapting control algorithm and self-rectifying control algorithm in some measure can solve the problem of no-certainty, however, from innate characters they still need to identify object model on line and thus lead to complicated algorithm and great computing. So they can not make us satisfied. For these reasons, in this paper we draw predictive control algorithm into the industrial furnace temperature control. We build the predictive model by simulation analysis. Then we proceed multi-step predication based on rolling choosing and feed-back correcting in order to obtain the newest information about industrial object. By relativity analysis, we know the output variable can trace the given value very well after application of predictive control algorithm, at the same time, the control system can resist disturbing rapidly. Consequently, it can meet the need of control system-rapid response function and strong robust function.In this paper, we have also discussed DMC-PID serial control algorithm. The algorithm uses different sample period at the different level in order to satisfy the need of choosing computer and resisting disturbing. In the inner loop, we adopt PID control algorithm, and in the outer loop adopt DMC control algorithm .So the algorithm has the virtues of serial structure and predictive control algorithm.Except theory studies we design the hardware and software of the furnace control system, which based on structure of field-bus. In the meantime, we also design the monitoring software by Visual C++ programming software. The whole system is very friendly and convenient to user.In this paper, many simulation curves are given. After analyzing those curves we can draw a conclusion that the DMC algorithm can improve the control property of system greatly. It is a good algorithm. Once the method of furnace control based on predictive control algorithm is applied widely, produce efficiency and production quality will enhanced enormously.