With development of technologies, people are becoming more and more dependent on technologies, and the same is true for the Heating, Ventilation and Air-Conditioning (HVAC) systems too. In the last around two decades, enormous development has taken place in the field of sensor technology, and that has made it possible to control the HVAC systems at every stage.
HVAC systems are the biggest consumers of energy in any modern building. For smart buildings, technologies have evolved to improve energy efficiency of HVAC systems, but faults often occur. Due to the complex nature of large-scale HVAC systems used in buildings, diagnosing these faults can be challenging.
The diagnosis algorithm
With this backdrop, a team of researchers led by Professor Marios Polycarpou, Director of the KIOS Research and Innovation Center of Excellence, Cyprus, has developed a distributed sensor fault diagnosis algorithm, a sequence of well-defined computer-implementable instructions for detecting and isolating multiple sensor faults in large-scale HVAC systems in smart buildings.
Putting more light on the background of the development, Professor Polycarpou, said, “The operation of Heating, Ventilation and Air-Conditioning (HVAC) systems in our homes, work spaces and public indoor spaces are based on the use of feedback measurements from sensing devices to make adjustments for maintaining a desired temperature. The presence of faulty measurements disorients the system and may create uncomfortable indoor conditions and/or significantly waste energy.”
Areas of application
This study presents an algorithmic approach that can be applied either on existing Building Management Systems or on plug-in Internet-of-Things (IoT) – a system of physical computer devices that are interconnected via a network for collecting and sharing data – to notify the building’s users and operators about the presence of faulty measurements, as well as the location of any faulty sensors.
Detection and isolation of sensor faults
In this study, the authors modeled a large HVAC system consisting of 83 building zones as a network of smaller interconnected sub-systems, rather than using a global model that describes the HVAC system for the entire building. This simplified method not only makes the design of model-based fault diagnosis more feasible, but it is also scalable, allowing for other parts of the building to be incorporated into the network using a plug-and-play approach.
According to Polycarpou, the utilization of thermal models of the variation of temperature in HVAC equipment and building zones, in combination with the design of diagnostic algorithms implemented in a multi-agent framework – a self-organized system consisting of several intelligent agents that interact with each other to solve complex problems that would be difficult for them to solve singularly – enables the development of advanced methods for detecting and isolating sensor faults. “In this framework, a wireless smart sensor can communicate with its neighbouring sensors to enhance the fault diagnostic process in terms of reliability, robustness, sensitivity and scalability,” Polycarpou explained.
The goal of the study
Describing the goal of their study, Polycarpou said, “Our ultimate goal is to develop lifelong diagnostic systems for smart buildings, which are able to continuously monitor their operation over the lifetime of the buildings, to detect, diagnose and self-heal any faulty behaviour, and to be able to learn from their prior experiences, as well as from the experiences of diagnostic systems from other smart buildings.”
(The complete study can be found at: https://ieeexplore.ieee.org/abstract/document/9080610)