Designing a neuro-fuzzy PID controller based on smith predictor for heating system

The most important part of the heating, ventilation, and air conditioning technology is heating System. This part is used in smart buildings and provides the desired air quality and thermal comfort. The time delay and uncertainty in model parameters due to the several operation mode cause the main challenges in heating system control by the traditional PID approaches. To overcome these problems, this paper presents an intelligent PID algorithm combines the fuzzy logic and neural network method together and used it in Smith predictor structure. Hence, a fuzzy neural network PID controller based on Smith predictor is proposed in this paper for the heating system. By correction of the dynamic learning of neural network and fuzzy inference, PID parameters of the controller get their optimal values. Simulation results of the heating system illustrate that the performance of the fuzzy neural network PID controller based on Smith predictor in comparison to the other control structures has been greatly improved, with fast response, smallest overshoot and lowest rise and settling time.