Research On Intelligent Recognition Method Of Aeroengine Impactors Based On Acoustic Emission

As a core component that provides flight power,aero-engines work in the natural environment and are vulnerable to foreign objects.After the the impact,it is necessary to make a maintenance plan according to the degree of damage.In order to save costs and improve efficiency,it is necessary to identify the type of impacts through condition monitoring and intelligent identification,and take countermeasures according to the type of impactor.In view of the early damage of the local structure when the aero-engine is subjected to impact,this paper proposes the use of acoustic emission monitoring means for real-time monitoring of the impact signal.Based on the plate impact test and the impact test of a real-life first-class moving blade,combined with the intelligent identification methods such as support vector machine and neural network,the research on the identification of aero-engine impactor is carried out.And for the characteristics of the sample data,the parameters of the intelligent recognition model are optimized to improve the recognition accuracy of the impact type.It is found through experiments that the type of impactor can be identified to some extent by using the acoustic emission characteristic parameters and the time and frequency domain parameters of the signal.In the plate experiment,it is found that the signal characteristics of different impactors are different.The input matrix of the support vector machine is optimized by independent element analysis.After the cross-validation of parameters,the correct rate of impact object recognition can reach 80%.In the aero-engine experiment,it was found that due to the complicated structure,it is difficult to extract the independent components of the acoustic emission signal.Therefore,to optimize the important parameter of the support vector machine,the particle swarm optimization algorithm should be used,and the correct rate of the impact object recognition reaches 64.7%.In order to further improve the accuracy of impact object recognition,a neural network impact object recognition method based on radial basis function is proposed.The recognition accuracy rate can reach 98.1%,which can be used as the basis for the identification of aero-engine impactors.Finally,the comparison of the two methods is given,which provides data support for the identification of aero-engine impactors,provides a new idea for the optimization of intelligent identification methods,and provides a new idea for establishing an aviation engine online fault diagnosis cloud platform.