Purification And Shape Recognition Of Shaft Orbits Of Hydraulic Power Sets

In recent years, with the rapid development of China's hydroelectric utilities, more and more large-scale hydraulic power sets have been used. As the size and capacity of the hydraulic power sets increases, automation level raises, the possibility of faults is also increasing. The shape of shaft orbits of the hydraulic power sets is an important symptom to diagnose the fault.Firstly, the causes of vibration of hydraulic power sets are analyzed in this paper. It also discusses the theories about vibration faults and the features of the vibration signal. Secondly, two methods are introduced to purify shaft orbits. The first one is to use wavelet theory. The second one is to use harmonic wavelet theory. To get better results, it is necessary to choose an appropriate wavelet and an appropriate layers when the wavelet theory is applied. Harmonic wavelet has specific function form, and the algorithm is simple, so harmonic wavelet is used to purify shaft orbits in this paper. By comparison, the harmonic wavelet makes a better effect in the purification.Finally, the moment invariant is adopted to recognize the shape of shaft orbits. As Hu's moment invariant is not satisfied with scale invariant on discrete condition, an improved algorithm is proposed in this paper. The improved moment invariant has the invariability in translation, rotation and scale in variant condition. Using the improved moment invariant to recognize the shape of shaft orbits is feasible.