Identification of Engine Foreign Object Impact Based on Acoustic Emission And Radical Basis Function Neural Network

Abstract

Engine blades are unavoidably impacted by foreign objects. In maintenance, the impact type is a decisive factor. But it is difficult to identify the impactor accurately, so how to make the accurate identification becomes the serious problem. This work proposes a method of foreign object identification for engine based on acoustic emission, time-domain analysis, frequency domain analysis and radical basis function neural network. Impact type recognition ability of neural network was tested by using non-homologous samples. The result shows that average signal level, peak frequency and center frequency can be used to identify impact type, and it also shows that neural network recognition accuracy is high and fast convergence.