The smart building, as an application of the cyber-physical systems (CPSs), plays an important role in everyday lives of people. Thermal comfort and energy efficiency are primary goals for HVAC systems in smart buildings. Since the controllers of the HVACs heavily rely on data of sensors that are deployed in the buildings, temporary or permanent sensor faults may lead to increased energy consumption or decreased thermal comfort far below expectations. In this paper, we examine sensor data faults observed in the real-world sensor deployments, and their effects on thermal comfort and energy efficiency in multi-room buildings. The read-back and nearest neighbor monitoring approaches are proposed considering temporal and spatial correlations between data of sensors to mitigate the faults of interest. We adopt a model-based design methodology for the multi-room building as a CPS application and develop reusable system models in the MATLAB/Simulink environment. We conclude that aforementioned faults may significantly reduce energy efficiency and thermal comfort unless mitigated. The proposed approaches improved thermal comfort by up to 75% for the room where the faulty sensor was deployed and reduced total energy consumption by up to 38%.
To appear in the 12th IEEE International Conference on Embedded Software and Systems (ICESS), August 2015