Joint Wavelet Transform and Time Synchronous Averaging Source Separation Method and its Application on Aero-Engine
Jiqing Cong,Jianping Jing,Changmin Chen,Zezeng Dai,Jianhua Cheng
DOI: https://doi.org/10.1115/1.0002539v
2021-01-01
Abstract:Abstract The reliability and safety of aero-engine are often the decisive factors for the safe and reliable flight of commercial aircraft. Hence, the vibration source location and fault diagnosis of aero-engine are of prime importance to detect faults and carry out fast and effective maintenance in time. However, the vibration signals collected by the sensors arranged on the casing of the aero-engine are generally the mixed signals of the main vibration sources inside the engine, and the components are extremely complicated. Therefore, the vibration source identification is a big challenge for a fault diagnosis and health management of the engine. In order to separate the key vibration sources of rotating machinery such as aero-engine, a Joint Wavelet Transform and Time Synchronous Averaging based algorithm (JWTS) is proposed in this paper. Based on the fact that the fundamental frequency and its harmonic and sub-harmonic components are generally included in the vibration spectrum of shaft fault signal of rotating machinery, wavelet transform and time synchronous averaging algorithm are combined to extract them. The algorithm completes separating the main vibration sources with three steps. First, the source number and fundamental frequency of each source are estimated using the wavelet transform. Second, every source is extracted from each observed signal by the time synchronous averaging method. Time synchronous averaging method can effectively extract a signal of cycle and harmonic rotor components and can suppress noise. Third, the optimal estimation of each source is determined according to signal’s 2-norm. Since the extracted source with a larger energy is closer to the real source, and signal’s 2-norm is a good indicator of the signal energy. Hence, the key vibration sources related to rotary speeds of the engine are obtained separately. The method is verified by synthetic mixed signals first. Three periodic signals of different frequencies are used to simulate the vibration sources of the aeroengine. The fundamental, harmonic and sub-harmonic components of them, as well as Gaussian white noise, are randomly mixed. The results show that the JWTS algorithm can estimate the number of the main sources and can extract each source effectively. Then the method is demonstrated using vibration signals of a real aero-engine. The results indicate that the proposed JWTS method has extracted and located the main sources within the aero-engine, including sources from the low-pressure rotor, high-pressure rotor, combustion chamber and accessory. Therefore, the proposed method provides a new fault diagnosis technology for rotating machinery, especially for a real aero-engine.