Fine Frequency Division Analysis On Hydropower Station Unit And Powerhouse Vibration Signal

Hydropower unit vibration and its induced plant structure vibration pose a major problem that concerns the safe and stable operation of hydropower stations. Some large stations put into production have appeared some vibration problems in different degrees, which threatens the normal operation and staff safety. The modern turbine units may appear some new dynamic characteristics during its development towards high head. This trend may also bring more unknown factors which impact on the operation of the units. Study on the vibration regularity of the unit can provide reference for future hydropower stations’ design and operation, and is a valuable subject in the hydropower construction field.This paper takes a hydropower station with a high head and a long diversion tunnel as its background, and reaches its study purpose through real test and fine frequency division analysis. The main contents and conclusion include:(1) The features of the vibration signals are summarized through the analysis of the vibration source. The signals have a wide range in the frequency domain. Their vibration sources are clear and have a big difference in energy. According to these features, combining the theory of wavelet transform and wavelet packet transform, a method of fine frequency division is established.(2) Using the method above, this paper analyzes the vibration signals of the key test points by breaking them down into signals of relatively narrow frequency band, in order to explain the cause and regularity of the vibration. The result shows that the 0~1Hz frequency band takes up a major energy in most test points, especially in the longitudinal and vertical direction. The main cause of the vibration is the low frequency hydraulic vibration source.(3) On the basis of previous research, this paper studies the vibration feature of the water body in penstock and analyzes the possibility of water body resonance.(4) By summarizing the discriminant method of closely spaced modal and identifying the modal of the simulation signal, the method which contains inverse attenuation index window and fine frequency division is chosen as the suitable separation method for closely spaced modes. The modal of the hydropower house is identified using this method.