Power distribution network single-phase earth fault judgment method, system and equipment and storage medium

Abstract

The invention discloses a power distribution network single-phase earth fault judgment method, system and device and a storage medium. The method comprises the steps that zero-sequence currents of all feeder lines under different fault working conditions in a power distribution network are sampled; performing GAF conversion on the zero-sequence current of each feeder line to obtain a GASF feature map of the zero-sequence current of each feeder line; the transformed GASF feature maps are labeled respectively; inputting the labeled GASF feature map into a pre-trained power distribution network single-phase earth fault prediction model, and predicting probabilities of different labels; and selecting the label with the maximum probability as a final result of feeder line judgment. The feature information of the zero-sequence current sequence after the fault is extracted, and the method is simple and convenient; and the trained network can be repeatedly used under the same working condition and does not need to be trained again.

Description

technical field [0001] The invention belongs to the technical field of distribution network of electric power system, and in particular relates to a method, system, equipment and storage medium for single-phase ground fault discrimination of distribution network. Background technique [0002] Most of the neutral points of the 10kV distribution network are grounded through the arc suppression coil, that is, the small current grounding system. According to statistics, the probability of a single-phase ground fault in a low-current grounding system is as high as 80%. At this time, the system can continue to operate with the fault for 1 to 2 hours, because the fault current is relatively small at this time, and has little impact on the system power supply. However, due to the continuous development of the urban distribution network, the grid structure tends to be more complex, the number of cable lines gradually increases, and the operation mode changes more and more, resulting ...

AI Technical Summary

Problems solved by technology

[0004] 1) Fault feature extraction: Most of the basis functions of existing methods such as Fourier transform, wavelet transform, and S transform are fixed, resulting in weak feature representation capabilities, and in the extraction process, they are not adaptive and easy to obtain Feature components without actual physical meaning are not conducive to the construction of identification criteria

Although the empirical mode segmentation algorithm has adaptive characteristics, its decomposition is prone to the problems of mode aliasing and endpoint effect

[0005] 2) A single line selection method cannot be applied to various fault conditions and network structure changes. Due to the complex nonlinear relationship between the fault features actually extracted in the fault line selection and the line selection results, an accurate line selection method cannot be used. Mathematical model to describe, so the neural network can be used to describe this nonlinear relationship

Images