Research On Vibration Isolation And Identification Of Unit And Powerhouse In Hydropower Station

With the increasing of large hydropower stations,the single-machine capacity and the physical dimension continue to increase,the structural stiffness of the unit and the powerhouse is relatively reduced,and the vibration is becoming more and more intense.There are many factors that cause the vibration of the powerhouse of hydropower station,the main vibration sources are mechanical,hydraulic and electromagnetic.This paper relies on the prototype vibration data of an underground powerhouse of a large hydropower station,the vibration source of each part is researched by traditional signal analysis method and blind source separation method,the main contents and conclusions are summarized as follows:(1)The combined vibration test of unit and powerhouse structure was carried out for a large hydropower station,the vibration signals are processed by the traditional time domain and frequency domain analysis methods,and the primary and secondary vibration sources of each part are found out.(2)According to the coupling characteristics of vibration of hydropower unit and powerhouse,and the improved blind source separation method is used to separate and identify vibration signals.Firstly,filtering the noise of signals,and the singular value decomposition(SVD)method is used to analyze the principal components of vibration signals,the number of vibration sources in the signal is determined by combining the eigenvalue source number estimation method and the information theory criterion(BIC,AIC),Then,the joint approximation diagonalization of eigenmatrices(JADE)algorithm is used to obtain the separation matrix;Finally,the main vibration sources in vibration signals of the unit and powerhouse structure of hydropower station were separated successfully: vortex belt,unit rotation,volute flow uniformity etc.(3)The traditional blind source separation method can not solve blind source underdetermined problem,in view of this limitation,the combination of empirical mode decomposition(EEMD),source number estimation method based on eigenvalue and information theory and joint approximate diagonalization method are adopted to research the single channel vibration signal blind source separation of powerhouse structures in complex environment.And the sparse component algorithm is used to obtain the real estimation of the source signals,and then the contribution rate of the vibration sources to the vibration is calculated,which provides a new method for distinguishing the vibration sources of the powerhouse in the complex environment.