Study of Neuron Adaptive PID Controller in a Single-zone HVAC System

This paper addresses the modeling and control of heating, ventilating, and air-conditioning (HVAC) system while the system is operating under uncertainty and nonlinearity. In order to acquire better performance in the HVAC system, the neuron adaptive PID control was studied through comparing traditional PID control in simulation, which is a method for adaptively adjusting the PID parameters using neuron. The results illustrate that neuron PID controller has the capability of self-studying and self-adapting. Through the response of neuron PID control is slower than PID control, the output of neuron PID control first enter stable belt and the anti-interference to white noise is stronger than PID control, which is adaptable to the uncertainty factors. In order to simplify the formulation and, hence, to emphasize the feature of the control algorithm, only a single-zone HVAC system is considered. However, the control design can easily be extended to multi-zone HVAC system