Method and system for evaluating short-circuit resistance of transformer winding based on octangular geometric transformation

Abstract

The invention provides a method and system for evaluating the short-circuit resistance of a transformer winding based on octane geometric transformation, and belongs to the technical field of power equipment monitoring. The method comprises the following steps: carrying out octane geometric modal decomposition on acquired transformer vibration signals to obtain a plurality of single components, carrying out similarity test on all the single components by using a Pearson correlation coefficient, setting a threshold value, carrying out significance analysis, and combining the components to obtain octane geometric modal components; calculating kurtosis factors of the octangular geometric mode components, presetting default values to form a group of feature vectors, setting an evaluation standard, training the extreme learning machine, obtaining a winding short-circuit resistance evaluation model of the octangular geometric transformation extreme learning machine, and taking the feature vectors as the input of the model, thereby obtaining an evaluation result. The method is better in real-time performance, is higher in discrimination accuracy, and can effectively discriminate the potential safety hazard of the transformer in long-term operation.

Description

Technical field [0001] The present invention relates to the field of electricity equipment monitoring technology. Background technique [0002] The power transformer carries an important task of changed electricity in the power system. Its operating environment and working state are very complex, and the short-circuit current impact caused by force-resistance short-circuit faults often occurs, and the short circuit current will cause a huge electromagnetic force to be deformed. The winding shape has a certain accumulation effect. When its deformation is not large, the normal operation of the transformer is very small; when the deformation is accumulated to a certain threshold, even if the overcurrent will make the transformer mechanical failure or insulated breakdown, the caused power failure will bring huge Economic loss. This puts higher requirements for the anti-short circuit ability of the transformer windings. During the process of operation of the transformer, the vibration...

AI Technical Summary

Problems solved by technology

The existing vibration signal analysis techniques mainly include wavelet analysis, empirical mode decomposition method and ensemble empirical mode decomposition; wavelet analysis method: this method fits the original vibration signal by selecting the wavelet basis function to obtain the vibration signal Energy characteristics in different frequency bands, but the wavelet basis function of this method is difficult to adapt to the signal, it needs to be continuously selected, and the frequency resolution decreases as the frequency increases; Empirical mode decomposition method: This method first uses cubic spline interpolation Obtain the upper and lower envelopes of the signal, calculate the mean value of the upper and lower envelopes, and subtract the mean value from the original signal to form a new discrete sequence. If the sequence still has local maximum or minimum values, continue to repeat the above steps , until the mean value of the upper and lower envelopes is zero, so as to obtain several eigenmode functions, but this method has the problem of mode aliasing when there are many frequency components and high noise, and the decomposition effect is not ideal; Decomposition: This method is an improved empirical mode decomposition. By adding uniform Gaussian white noise before decomposition, it is expected to reduce mode aliasing. However, due to the addition of noise, this method will have false components after decomposition. Due to the addition of Noise, the calculation time is also relatively extended

Images