A single-phase-to-earth fault location method based on neural network and characteristic matrix

Abstract

The invention discloses a single-phase grounding fault location method based on a neural network and a feature matrix. The technical solution includes the following steps: Step 1, extracting the features of a data set collected on site, and using the extracted features as the input of the neural network; Step 2 , combined with the on-site topology map and the actual fault judgment results, add labels to the collected data set; step 3, normalize all the features, and divide the processed data set into training set and test set; step 4, Neural network model parameter tuning, training, testing, and output of fault vectors; step 5, establish node feature matrix through the site topology map, and combine fault vectors to obtain fault information feature matrix; step 6, divide fault sections, calculate fault metrics, and determine fault areas section position. The present invention is simple to implement, without adding other hardware devices, and can accurately obtain the position of the section where the fault point is located only by locally judging the voltage and current data measured by the fault indicator.

Description

technical field [0001] The invention relates to the technical field of medium-voltage distribution network, in particular to a single-phase ground fault location method based on neural network and feature matrix. Background technique [0002] The operating status of the distribution network is crucial to the stable operation of the power grid. Single-phase ground faults are the most common faults in power distribution systems. Single-phase ground faults not only affect the normal power supply of users, but also may generate overvoltage, which will cause The equipment is damaged, and even causes a short circuit between phases and expands the accident. However, there are a large number of terminals in the medium-voltage distribution network, various grounding methods, and complex data. It is difficult to locate accurately only by a single fault diagnosis method. Positioning poses enormous difficulties. Contents of the invention [0003] Aiming at the deficiencies and defec...

AI Technical Summary

Problems solved by technology

[0002] The operating status of the distribution network is crucial to the stable operation of the power grid. Single-phase ground faults are the most common faults in power distribution systems. Single-phase ground faults not only affect the normal power supply of users, but also may generate overvoltage, which will cause The equipment is damaged, and even causes a phase-to-phase short circuit and expands the accident

However, there are a large number of terminals in the medium-voltage distribution network, various grounding methods, and complex data. It is difficult to locate accurately only by a single fault diagnosis method. Positioning poses enormous difficulties

Images