Experimental studies of air-handling units’ faulty operation for the development of data-driven fault detection and diagnosis tools: A systematic review

Automated Fault Detection and Diagnosis (AFDD) algorithms could represent one of the most effective solutions in order to reduce energy demand, greenhouse gas emissions and running costs of Heating, Ventilation and Air-Conditioning (HVAC) systems equipped with Air-Handling Units (AHUs). In particular, data-driven AFDD tools are recognized as easier to be developed and able to provide a higher accuracy with respect to other AFDD tools. However, they are still in the early stage of adoption stock-wide mainly due to the facts that data-driven AFDD models (i) require labelled labeled and reliable experimental faulty data that are time-consuming and expensive to be obtained under different operating scenarios, and (ii) cannot operate beyond the training data. In this paper the most significant scientific papers focusing on experimental analyses of AHUs aiming at the development of data-driven AFDD algorithms have been systematically reviewed and categorized in order to highlight the most important research gaps to be still covered. In particular, the AHU operating schemes, fault types, faults severities and climatic conditions requiring further studies have been identified with the main aim of supporting and guide the future development of new and accurate data-driven AFDD systems.