Detection of performance degradation in aircraft engines with low pressure turbine vibration parameter by using artificial neural network and fuzzy logic

Abstract

Sensors in the aircraft are used to receive information about the aircraft and its components. The pilot is controlling the aircraft by monitoring the data from sensors on displays. The data from the sensors are evaluated in the engine health monitoring (EHM) and condition monitoring system to ensure that we can identify faults that may occur. Performance changes in aircrafts are monitored by parameters. Vibration analysis is an effective method of diagnosing faults and ensures that faults in the mechanical construction of the aircraft can be detected easily. Vibration is the response of a machine’s mechanical components to internal and external forces. In this study, in the first step, the effect of flight parameters on the Low Pressure Turbine (LPT) vibration parameter was investigated by multiple regression analysis (MRA). According to the results of the analysis, Thrust Lever Angle Right, Airspeed, EGT Left and N2 Speed Left parameters of the flight parameters were found to be important parameters effecting LPT vibration parameters. The second step is modeling the artificial neural network (ANN) which is developed in the MATLAB SIMULINK environment, to estimate an LPT vibration parameter. 62 data samples recorded during the flight of an aircraft. For this model 44 of 62 data samples were used in the training, 9 of 62 samples were used for test and 9 for validation of models. In the last step, engine performance is evaluated with the fuzzy logic (FL) interface developed to compare the LPT vibration value obtained from the sensor of the aircraft with the estimated LPT vibration value. With this study, pilots will be able to obtain more detailed information about the engine health condition during flight.