Application of classification functions to chiller fault detection and diagnosis

This paper describes the application of a statistical pattern recognition algorithm (SPRA) to fault detection and diagnosis of commercial reciprocating chillers. The developed fault detection and diagnosis module has been trained to recognize Jive distinct conditions, namely, normal operation, refrigerant leak, restriction in the liquid refrigerant line, and restrictions in the water circuits of the evaporator and condenser. The algorithm used in the development is described, and the results of its application to an experimental test bench are discussed Experimental results show that the SPRA provides an effective way of classifying patterns in multivariable, multiclass problems without having to explicitly use a rule-based system.