Optimal adaptive predictive control and fault detection of residential building heating systems

Abstract

We develop an optimal predictive control for building heating, ventilating and air conditioning (HVAC) systems. Explicit inequality constraints on the input and on the output of the system are considered. A specific model-based recursive parameter identification and fault detection approach is described. Simplified physical modeling of the process establishes the model structure of an adaptive predictor to be used both for identification and control. The adaptive predictor is based on the estimation of the states. The parameters of the process are updated using a two-stage filtered instrumental projection method that ensures the promptness of fault detection. Fault detection is based on specific processing of the prediction errors and parameters of the predictor. Process diagnosis is ensured by appropriate use of the qualitative knowledge about the process. In order to increase the robustness of the fault detection scheme, additional test signals are introduced in the process.