Author:
Yu Shige,Zeng Yun,Dao Fang,Li Xiang,Xu Yiting,Qian Jing
Abstract
Abstract
In recent years, the research on characteristic extraction methods for non-stationary and non-linear vibration signals of hydroelectric units has been a hot topic in the fault diagnosis of hydroelectric units. The effectiveness of feature extraction directly affects the accuracy of fault diagnosis. This article proposes a vibration signal feature extraction method for hydroelectric units based on motion model separation and logarithmic attenuation rate method. The under-damped decay oscillation motion trajectory is obtained from the signal after motion model separation. And then, the logarithmic attenuation rate method in the time domain analysis is used to calculate the system damping ratio and natural frequency. In this process, the principle of selecting the envelope interpolation function is proposed, and the effectiveness of the cubic Hermite interpolation function is determined. Based on the calculation effectiveness and accuracy requirements, the effective step length of the vibration signal is determined to be Δt = 0.001, and the effective calculation interval for extracting the decay oscillation characteristics is [ts
, tend
]. The results show that the proposed method is more straightforward, efficient, has higher computational accuracy and better signal-to-noise ratio adaptability, and can provide a theoretical basis for practical engineering applications.
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献