Trend prediction of wear fault of wind generator high-speed gear using a fusion of UICA and PE method

The large wind generating set works under the varying operation conditions for years, generally the fault characteristic values of the rotary components based on the energy mode is coupled with other noises, so the fault trend can not be accurately predicted. With the wear fault of the high-speed gear of the wind generator as the research object, this paper proposes the trend prediction method of the wear fault of the high-speed gear based on the fusion of ultra-complete independent component analysis (UICA) and parameter estimation (PE), constructs the ultra-complete analysis model, separates the similar source signals more than mixing signals by using the UICA, and finds useful component with the features of the pure similar fault source signal as the basis. Based on the similarity between the similar fault source signal and fault source signal, this paper estimates the value domain of the magnification time of the similar shapes by using the PE, identifies the mapping between continuous and one-way varying magnification time domain and rotary component fault degree, establishes the fault degree judgment standard, and determines and predicts the fault degree and the fault trend based on the energy change trend diagram of the whole-lifecycle fault source signal of the high-speed gear. The above method is used to process the wear fault data of the high-speed gear. The results indicate that the above method has obvious effect in processing of the cycle sudden signals, so it indicates that this method has certain engineering application value and provides reference to solve the problem that the number of the independent vibration sources is more than it of the mixing signals in vibration analysis.