A LITERATURE REVIEW OF AUTOMATED FAULT DETECTION AND DIAGNOSTICS FOR HVAC SYSTEMS

Modern HVAC systems are significant energy consumers and major contributors to environmental pollution. These systems consume around 52% of the total U.S. energy used in commercial buildings on average; withal, the air conditioning industry is responsible for approximately 4% of global greenhouse gas emissions annually. Previous research has targeted ways to improve the energy efficiency of HVAC systems and reduce their environmental impact via enhanced control algorithms. However, because maintenance teams execute HVAC system diagnostics offline, this hinders detecting and correcting any system malfunctions or underperformance immediately. Currently, innumerable research is underway that focuses on improving automated fault detection and diagnostics (AFDD) algorithms for HVAC systems to reduce operating costs and improve efficiency and indoor air quality (IAQ). This study aims to synopsize the typical fault detection and diagnosis systems being developed for commercial and industrial buildings. Still, the authors believe there is a need for less expensive AFDD algorithms for HVAC systems, as well as common standards or protocols for FDD development. Machine learning may be the best approach to achieve a more dynamic system that adapts to changing situations independently, which could be the key to developing an adaptive HVAC AFDD system using Artificial Intelligence.