Search

This study conducted modal and vibration-characteristic tests using transient excitation and spectrum analysis methods to find the variation law of vibration characteristics of wind wheels with bifurcated tip structures as compared to wind wheels with unmodified tips when the coupling between a wind wheel and its tower is considered. Additionally, a finite element analysis was used to calculate the mechanical characteristics of wind power manoeuvres. The following are the major results of this study. As compared to the non-trailer state, the coupling action of the wind wheel and the tower reduced the static frequency of the unmodified wind wheel and the bifurcated blade tip structure wind wheel under the trailer state. The static frequency of the bifurcated blade tip structure wind wheel under the coupling action decreases more significantly. Compared with the single wind wheel which is not affected by coupling, the dynamic frequency of the whole machine decreases after being affected by coupling, and the bifurcated blade structure has less influence on the coupling effect. Compared with the non-trailer state, the dynamic frequency curves of the two blade tip structure wind wheels in the trailer state decrease, the speed range in the resonance area is shortened, and the corresponding speed in the resonance area is reduced. The results of this study provide data support and design reference for reliable design of wind turbines.

This study aims to solve the problem of extreme point ambiguity caused by energy instability at the signal end. Thus, an adaptive nonlinear signal decomposition method based on motion energy accumulation division is proposed, namely slope integral extension mode decomposition (SIEMD). The proposed method considers the fluctuation rate and vibration energy between the peaks of the waveform as its scale. Firstly, the comprehensive index is defined to adaptively select the ideal interval, and the extension characteristics of the waveform signal are obtained. Secondly, the energy of the waveform interval is iterated. Hence, the optimal extension waveform is fitted by combining the edge position information of the curve. The experimental part verifies that the method can extract 92 % of the fault information, and verifies that the proposed method overcomes the limitation of the previous one-dimensional signal waveform dimension. Moreover, from the perspective of signal energy, it eliminates the false information of the intrinsic modal function (IMF) components, more suitable for the randomness of the signal, thereby providing a new way for fault feature extraction.

NPP (Neighborhood Preserving Projections) is an incremental subspace learning methods which has a nature of maintaining the data local neighborhood geometry constant. To improve the discriminatory power of NPP, NPD (Neighborhood Preserving Discrimination) algorithm was proposed to be used for the rotor system fault data set feature dimensionality reduction. Floyd algorithm based on graph theory and MMC (Maximum Margin Criterion) were introduced in the NPP which makes NPD avoid the short-circuit problem that occurs in the high curvature high dimensional space data sets, while enhancing data discrimination information during the dimensionality reduction. In addition, NPD can maintain the manifold of data set unchanged. At last, the rotor-bearing experiment has been made to verify the effectiveness of the NPD method.