Coupling a neural network temperature predictor and a fuzzy logic controller to perform thermal comfort regulation in an office building

The paper describes the application of a combined neuro-fuzzy model for indoor temperature dynamic and automatic regulation. The neural module of the model, an auto-regressive neural network with external inputs (NNARX), produces indoor temperature forecasts that are used to feed a fuzzy logic control unit that simulates switching the heating, ventilation and air conditioning (HVAC) system on and off and regulating the inlet air speed. To generate an indoor temperature forecast, the NNARX module uses weather parameters (e.g., outdoor temperature, air relative humidity and wind speed) and the indoor temperature recorded in previous time steps as regressors. In its current state, the fuzzy controller is only driven by the indoor temperature forecasted by the NNARX module; no variations in indoor heat gains or occupantsʹ clothing and behavior were considered for driving the controller. The main goal of this paper is to demonstrate the effectiveness of the hybrid neuro-fuzzy approach and the importance of efficiently designing the temperature forecast model, especially with respect to the selection of the order of the regressor for each of the external and internal parameters used. Therefore, a differential entropy-based method was applied in this study, which provided good forecasting performances for the NNARX model.