Study On Stability Monitoring And Operating Performance Analytical Methods Of Hydroelectric Generating Set

With the hydropower development, Hydropower unit’s capacity and scale is growing,and the stability of operating has become more and more important of hydroelectric powerset. The state overhaul is also beginning to develop step by step, the methods of faultdiagnosis are more and more proposed. This paper has developed stability monitoringsystem of hydropower units to grasp the operation of the unit state in time, which canprovide technical support for the intelligent diagnosis.This paper discusses how to develop the stability monitoring system by Visual C++6.0using C/S mode, it can achieve data sharing in the local area network. The system hasthe function of real-time display waveform as well as the characteristic value, and caninquire the detail information of failure, plays an important role on improving the safetyand economic behavior of the hydropower plant.The identification method of axis orbit based on PSO-BP was proposed. This methodis based on theoretical basis of information fusion, Combination of moment invariantsserves as a characteristic vector of an axis orbit of a hydro turbine, which is composed byimproving Hu invariant moment and affine invariant moment, and achieves the goal ofaxis orbit identification using BP neural network optimized by PSO. It can provideguidance of finding out the fault of hydroelectric units.A method of analyzing hydro-generator unit vibration signals based on EMD-FFTwas proposed. This method is a new method of time-frequency analysis which is thefusion of FFT and the empirical mode decomposition (EMD). First the original signal isdecomposed adaptively and steady using time-frequency focusing of EMD, thendistinguish the signal mutation moment from each IMF, and acquire the time-frequency-amplitude by FFT.