Research On Feature Recognition Method Of Common Faults Of Aero-engine AE Signal Based On Rough Set

Aircraft engine is the ’heart’ of the aircraft,and is the extremely complex mechanical systems,and played a vital role to the safe operation of the aircraft.The faults of the bearing and rotor system are the most common,such as the blade fracture,the impact of the foreign body,etc.In the bearing failure,the inner and outer wear of the bearing and the rolling off of the rolling body often occur.The rolling bearing is the main support components of the aero engine,therefore,in recent years,there are more and more extensive research on rolling bearings.In this paper,acoustic emission technology is used as the primary means.Under the condition of acoustic emission technology,the fault bearing signal is analyzed and the theoretical and experimental researches are carried out.Mainly for three types of failure:inner ring,outer ring and rolling body,specifically to do the following aspects of the study:The acoustic emission and the theoretical knowledge related to the acoustic emission are introduced.The propagation theory of the acoustic emission wave,the time domain of the acoustic emission signal,the frequency domain and the wavelet packet are analyzed.And briefly introduce the characteristics and application of acoustic emission technology.Then,built a simple engine test bench,acoustic emission signals in normal and three fault conditions are measured.The characteristics of the acoustic emission signal are analyzed from the aspects of acoustic emission characteristic parameters,time domain analysis and frequency domain analysis.The faults of the bearings are simply identified,but the fault types can’t be accurately distinguished.Lastly,based on the artificial intelligence algorithm-neural network,a fault diagnosis method based on rough set and neural network is proposed.The feature data is extracted by wavelet packet analysis,and the attribute is obtained by rough set.Input the reduced data to the neural network to complete the fault type of identification,greatly improved the fault recognition rate.And prove the effectiveness of this method.In this paper,the fault identification and diagnosis method of the rolling bearing of aero-engine is presented,which can improve the fault recognition rate and identify the fault types.It provides a guarantee for the safe operation of aero-engine.