Research On Precision Temperature Control System Using Fuzzy Neural Networks

Temperature is a common and one of the most basic process variables in industrial production, scientific research and daily life. Temperature control exists widely in every aspect of social life, and PID is the mainly adopted control algorithm in practice. Due to the environment variation PID parameters in-process tuning is not attainable, which negatively effects control quality.Control theory is ceaselessly updated with the development of science and technology. Fuzzy Control and ANN are newly emerging control technologies. This paper explores the method to combine Fuzzy control and ANN, and applies it to precision temperature control. The structure, algorithms, and applications of Fuzzy control and ANN control system is systematically studied.Taking the oil cooler heat transfer characteristics measurement system as the research object, this paper present the hardware design method, and uses Visual C on the windows as the tool to implement the Fuzzy ANN control algorithm. This software also implements temperature real-time control, date collection, analysis and processing, and automatic drawing as well as print. This testing system, with high reliability, has been successfully applied to real-life production. It is practically proved that this method is independent with the plant mathematic model, has small over shoot, short adjusting time, and good robustness, and deserving promotion.