Fuzzy Predictive Control Technology Applied Research In The Lag Temperature System

Time delay is a common problem in industry. There are many large time-delay systems is industry. Because of time delay, disturbs can not be detected in time and it takes long time to affect the objects under control. These influence controlling precision, static error and stability remarkably.Fuzzy control is an apery technology which is independent of mathematic model. It uses experts' experience and knowledge to build the fuzzy rules. Traditional fuzzy control is a post pose adjust technology which is similar to PID. The control effect for large delay object is not ideal. Predictive control focuses on the future target as well as past and current target. Predictive control and fuzzy control are two independent control methods. The combination of the two methods is proposed to form a new control method, that is, fuzzy predictive control.In this paper, the property of large time delay objects and evolvement in and aboard is introduced firstly. Then the implement of fuzzy control, predictive control and fuzzy predictive control is presented. For exigent circumstance, model identification is all important. In the third chapter, experiments are designed for model identification of calibration furnace. The piecewise modeling scheme is proposed finally. For the time-delay and non-linear property of calibration furnace temperature control system, different fuzzy rules sets for different piecewise are designed to improve control effect. And predictive control method is used to solve the large time-delay problem. One-order pure time-delay model is simulated using fuzzy predictive control scheme. And the scheme is used to control the calibration furnace. It is shown that fuzzy predictive control technology has the merit of two control methods. The system is robust and accurate.Temperature control system of calibration furnace is a typical non-linear large time-delay system. Fuzzy predictive control realizes effective smooth control of calibration furnace. The static error is less than 0.5℃. the temperature variance is less than 0.2℃/min. the steady-state time is more than 30 minutes. The control scheme and its effects provide reference for other large time-delay process.