A process-history based fault detection and diagnosis for VAVAC systems

Faulty operations of Heating, Ventilation and Air Conditioning (HVAC) systems can lead to discomfort for the occupants, energy wastage, unreliability and shorter equipment life. Cost-effective Fault Detection and Diagnosis (FDD) methods can therefore ensure an increase in the system uptime, reliability, and overall efficiency. In this paper, a simulation environment based on Matlab/Simulink R is used in order to evaluate the performance of a FDD method using Support Vector Machines (SVMs). In detail, the proposed method is evaluated by performing extensive simulations to allow the investigation of the most common and relevant faults affecting this kind of systems.