Presents an artificial neural network (ANN) prototype for fault detection and diagnosis in complex heating systems. States the six operating modes with faults used to develop the prototype stemmed from a detailed investigation in cooperation with heating systems maintenance experts and are among the most important operating faults for this type of system. Demonstrates the feasibility of using ANNs for detecting and diagnosing faults in heating systems provided that training data representative of the behaviour of the systems with and without faults are available.

KEYWORDS: year 1997, Heating, failure, expert systems, detectors, feasibility