Electric Power ›› 2019, Vol. 52 ›› Issue (9): 154-160.DOI: 10.11930/j.issn.1004-9649.201903020

Previous Articles     Next Articles

Noise Analysis and Signal Processing in the Turbine Valve Closing Test

MEI Yiming1, GU Weifei1, ZHU Xiangyun2, ZHU Bao1   

  1. 1. Zhejiang Key Laboratory of Energy Efficiency and Pollutant Control Technology for Thermal Power Generation, Hangzhou 311100, China;
    2. Zhejiang Energy Technology Research Institute Co., Ltd., Hangzhou 311100, China
  • Received:2019-03-18 Revised:2019-05-20 Online:2019-09-05 Published:2019-09-19

Abstract: The valve position signal may be contaminated by noise during the turbine valve closing test, which makes it hard to calculate the closing time accurately. As a result the reliability of turbine safety assessment will seriously be affected. In this paper, the components of the contaminated signal sampled in turbine valve closing test were first analyzed using fast Fourie transform (FFT). Then the accurate closing time was calculated by waveform comparison regarding the noise signal with single sinusoidal waveform component. Afterwards, the signal contaminated by multi-frequency sinusoidal noises was filtered by infinite impulse response (ⅡR) digital filtering method. The filtering effects of different filters were compared and the closing time after the filtering were calculated. It is suggested the waveform comparison method is very suitable for the computation of closing time with the signal contaminated by single component sinusoidal waveform. Both the Butterworth filter and Chebyshev I type filter can be used for signals contaminated by multi-frequency sinusoidal noises, while the former is slightly better than the latter regarding the accuracy of calculation results.

Key words: steam turbine, turbine valve closing time, noise contamination, FFT, waveform comparison, ⅡR digital filtering

CLC Number: