Hydropower is the green energy with the most significant comprehensive emission reduction benefits in the whole life cycle. The signals of hydropower equipment include fault information, and it can assist the fault diagnosis of hydropower units. However, most of the existing methods lack the quantitative evaluation of the equipment degradation. A quantitative evaluation method of degradation for hydropower equipment is proposed. Variable modal decomposition (VMD) is employed to obtain decomposed simple signal. The singular values and sample entropy of the intrinsic mode functions are obtained and combined into a feature vector. Jenson-Shannon divergence is adopted to evaluate the degradation of hydropower equipment by comparing the current feature vector with the normal state feature vector. Experimental results show that this method can provide degradation evaluation information. The proposed method can provide not only quantitative indicators of equipment degradation, but also early warning of equipment degradation than the usual anomaly detection methods.
本站僅提供期刊文獻檢索。
