RARX algorithm based model development and application to real time data for on-line fault detection in Vav Ahu Units

Assimilation of cost-effective Fault Detection and Diagnosis (FDD) technique in building management sys- tem can save enormous amount of energy and material. In this paper, Recursive Autoregressive Exogenous Al- gorithm is used to develop dynamic FDD model for vari- able air volume air handling units. A methodology, based upon frequency response of the model is evolved for automatic fault detection and diagnosis. Results are validated with data obtained from a real building after introducing artificial faults. It is concluded that the method is quite robust and can detect and diagnose several types of faults