Anomaly detection in aircraft gas turbine engines

The application of a novel method for early detection of anomalies in complex mechanical systems is discussed. The anomaly detection method is based on symbolic time series analysis (STSA) and its efficacy has been examined on the simulation test bed of a twin-engine propulsion system. the time series data of observed macroscopic variables, generated on the first time scale from the simulation model, are analyzed at slow time scale epochs for each detection of anomalies. the method is found to be more efficient compared to principal component analysis (PCA) and artificial neural network (ANN) methods.