DENSITY-BASED CLUSTERING ALGORITHM FOR FAULT DETECTION AND IDENTIFICATION IN HVAC SYSTEMS

The operation of building services like heating, ventilation and air conditioning systems (HVAC systems)is often vitiated by faults and sub optimal states. Such malfunctioning can be overcome by introducing a monitoring system with automated fault detection and diagnosis based on measurement data.Here, we propose a method for the automatic detection and identification of faults in HVAC systems,which is based on a clustering algorithm. We illustrate the proposed method using simulation data from a simple model, where faults have been implemented artificially, and show that the approach performs well with respect to fault detection and that it provides additional valuable information to enable fault diagnostics.