Application of wavelet time-frequency analysis on fault diagnosis for steam turbine

Gang Zhao,Dongxiang Jiang,Jinghui Diao,Lijun Qian
2004-01-01
Abstract:Many vibration signals of fault turbine rotor are nonstationary and have highly complex time-frequency characteristics. A good way to analyze the nonstationary mechanical signals is the wavelet transform. Unfortunately, the direct result of wavelet transform in fault diagnosis of steam turbine is not easy to understand because it lack of physical meaning compared with FFT. Consequently, time-frequency contour map is introduced into fault diagnosis of steam turbine. The time-frequency contour map can easily show the power distribution of signal in time and frequency domain. Moreover, it is a good way to identify the steam turbine faults involving a breakdown change. Six typical faults of steam turbine are discussed by the time-frequency contour map in the paper. The simulative results show that time-frequency contour map have the capabilities to identify the difference of those faults of steam turbine. In conclusion, the faults of steam turbine can be classified by time-frequency contour map. Key word: time-frequency analysis, wavelet, fault diagnosis, steam turbine
What problem does this paper attempt to address?