Dr. Goran Andonovski, univ. dipl. inž., Univerza v Ljubljani, Fakulteta za elektrotehniko; dr. Aljaž Stare, univ. dipl. inž., Metronik, d. o. o., Ljubljana; prof. dr. Igor Škrjanc, univ. dipl. inž., Univerza v Ljubljani, Fakulteta za elektrotehniko

Abstract:
Fault detection in industrial systems plays a crucial part of the correct and optimal operation of the different processes. The possible faults could be of different types and natures and occur on different components of the process. This paper presents fault detection on air conditioning systems using an advanced data-based method. The method evolves its structure in an online manner and builds various fault detectors for each fault. Furthermore, we could detect and identify each fault separately. The proposed method was tested on real data from a real air conditioning system. The advantage of the evolving method is the ability to detect new faults before it affects the operation of the process.

Keywords:
advanced methods, fault detection, HVAC system