Combining knowledge and historical data for system-level fault diagnosis of HVAC systems

Abstract

Interdependencies among system components and the existence of multiple operating modes present a challenge for fault diagnosis of Heating, Ventilation, and Air Conditioning (HVAC) systems. Reliable and timely diagnosis can only be ensured when it is performed in all operating modes, and at the system level, rather than at the level of the individual components. Nevertheless, almost no HVAC fault diagnosis methods that satisfy these requirements are described in literature. In this paper, we propose a multiple-model approach to system-level HVAC fault diagnosis that takes component interdependencies and multiple operating modes into account. For each operating mode, a distinct Bayesian network (diagnostic model) is defined at the system level. The models are constructed based on knowledge regarding component interdependencies and conservation laws, and based on historical data through the use of virtual sensors. We show that component interdependencies provide useful features for fault diagnosis. Incorporating these features results in better diagnosis results, especially when only a few monitoring signals are available. Simulations demonstrate the performance of the proposed method: faults are timely and correctly diagnosed, provided that the faults result in observable behavior.

Fault Detection and Diagnostics in Ventilation Units Using Linear Regression Virtual Sensors

“…They develop Bayesian networks for multiple operating modes, using both actual and virtual sensors created from system knowledge and historical data. The authors show how using virtual sensors significantly improves FDD performance [13].…”

A Method for Fault Detection and Diagnostics in Ventilation Units Using Virtual Sensors

“…In [ 22 ] the authors report how using virtual sensors significantly improves FDD performance for HVAC systems. They propose a multi-model FDD method that exploits components interdependencies.…”

“…The reviewed state of the art shows that virtual sensors are popular in the field of buildings systems; however, to our knowledge there is no work so far on employing data-driven virtual sensors for fault detection and diagnostics application on ventilation units. Most of the work reviewed covers other buildings subsystems, such as chillers and air-conditioning units [ 20 , 23 , 24 ], boilers [ 22 ], heat pumps [ 20 ] and room-level components [ 25 ]. Moreover, in ventilation units, virtual sensors are usually developed to provide readings for unmeasured quantities [ 21 ], and when they are considered for explicit application for fault detection and diagnostics they are designed using first principles methods [ 20 ].…”

“…Moreover, in ventilation units, virtual sensors are usually developed to provide readings for unmeasured quantities [ 21 ], and when they are considered for explicit application for fault detection and diagnostics they are designed using first principles methods [ 20 ]. Other approaches focus on a higher level of diagnostics and require significant expert knowledge to define fault and symptoms [ 22 ]. Therefore, the main contribution of this paper is a specific fault detection and diagnostics application for ventilation units based on virtual sensors created using a data-driven approach .…”

A review of model based and data driven methods targeting hardware systems diagnostics

“…In [114] there is developed a system level diagnostic approach applied on a HVAC system. The proposed method considers the system components interdependencies.…”