Intelligent system based on wavelet decomposition and neural network for predicting of fan speed for energy saving in HVAC system

In this study, a heating, ventilating and air-conditioning (HVAC) system with different zones was designed and tested. Its fan motor speed and damper gap rates were controlled by two controllers (i.e. a PID controller and an intelligent controller) in real time to minimize its energy consumption. The desired temperatures were realized by variable flow-rate by considering the ambient temperature for each zone and evaporator. The PID parameters obtained in our previous theoretical work using fuzzy logic were utilized in this study. The experimental data used in this study was collected using a HVAC system built in a laboratory environment. The fan motor speed and damper gap rates were predicted using wavelet packet decomposition (WPD), entropy, and neural network (NN) techniques. WPD was used to reduce the input vector dimensions of the intelligent model. The suitable architecture of the NN model is determined after certain trial and error steps. According to test results, the developed model performance is at desirable level. Efficiency of the developed method was tested and a mean 95.62% recognition success was obtained. This model is an efficient and robust tool to predict damper gap rates and fan motor speed to minimize energy consumption of the HVAC system.