Fuzzy Model Predictive Control for a Comfort Air-Conditioning System

Predictive control strategy is presented for a comfort air-conditioning system concerning energy performance and indoor comfort requirements. Predicted mean vote (PMV) index was incorporated into the control strategy to assess thermal comfort level. First-order linear reference trajectory was selected for the controlled variable PMV. And a zero-order Takagi-Sugeno fuzzy model, being linearized in real-time around the current operating point, was adopted as the predictive model. The experiment results indicate that the proposed techniques can save energy and improve indoor comfort significantly for air conditioning systems compared to the conventional PID control technique.