A fault diagnosis approach for airborne fuel pump based on EMD and probabilistic neural networks

Abstract

Airborne fuel pump is an important part of aircraft fuel system, for the lack of fault samples, low diagnostic efficiency and high maintenance cost, in order to achieve more accurate and reliable fault diagnosis of airborne fuel pump, an experimental platform of fuel transfusion system is developed and a fault diagnosis method based on empirical mode decomposition (EMD) and probabilistic neural networks (PNN) is proposed. Firstly, the vibration signals and pressure signals of normal state and six types of representative fuel pump faults are collected on the experimental platform. Then the EMD method is applied to decompose the original vibration signals into a finite Intrinsic Mode Functions (IMFs) and a residual. Secondly, the energy of first four IMFs is extracted as a vibration signals failure feature, combined with the mean outlet pressure to structure the fault feature vector and then divided into training samples and testing samples. Training samples are used to train the PNN fault diagnosis model and testing samples are used to verify the model. Experimental results show that compared with SVM and GA-BP, the PNN fault diagnosis model has fast convergence, high efficiency and a higher performance and recognition for the typical faults of airborne fuel pump.