Abstract
The objective of this work is to present a method to monitor the health of Auxiliary Power Units (APU) using a Dynamic Computational Model, Gas Path Analysis and Classification and Regression Trees (CART). The main data used to train the CART consists of measurements of the exhaust gas temperature, the bleed pressure and the fuel flow. The proposed method was tested using actual APU data collected from a prototype aircraft. The method succeeded in classifying several relevant fault conditions. The few misclassification errors were found to be due to the insufficiency of the information content of the measurement data.