Deep learning-based distributed power supply-containing power distribution network fault diagnosis method

Abstract

The invention belongs to the technical field of power grid fault diagnosis, and particularly relates to a deep learning-based distributed power supply-containing power distribution network fault diagnosis method, which comprises the following steps of: generating a sample set, wherein data is acquired by adopting an FTU, and preprocessing the acquired real-time data; obtaining a fault type based on the preprocessed data; inputting a distributed power supply-containing power distribution network before and after a fault into a neural network corresponding to the fault type to obtain a line number of a fault occurrence place; inputting the preprocessed data into the neural network corresponding to the fault type to obtain a first line location of a fault location in a power distribution network line; performing secondary refined fault positioning by combining a fault grading processing method, and screening non-fault areas; and quickly obtaining a second short-circuit position based on an improved particle swarm algorithm of adaptive chaos variation, comparing the second short-circuit position with the first line location, if the second short-circuit position and the first line location are consistent, obtaining a final fault diagnosis result, and if the two are not consistent, repeating the above steps.

Description

technical field [0001] The invention belongs to the technical field of power grid fault diagnosis, and in particular relates to a method for diagnosing faults in distribution networks containing distributed power sources based on deep learning. Background technique [0002] The operating status of the distribution network is directly related to the reliability and power quality. The public power households have gradually increased their requirements for reliability and power quality. However, in the actual operation of my country's distribution network, its power supply reliability is not high, and the average power outage time is longer than that in Europe and the United States. The level of developed countries is mainly due to the short circuit problem in the distribution network. Short-circuit faults in the distribution network have had a great impact on industrial production and residents' lives. Quickly locating and removing faults in the distribution network can effecti...

AI Technical Summary

Problems solved by technology

[0005] Aiming at the problem that the traditional fault location method is difficult to satisfy the distribution network with distributed power supply, this invention proposes a fault diagnosis method for distribution network with distributed power supply based on deep learning, and uses the advantages of deep learning in approximation ability and fault tolerance to dig The mapping relationship between the response data and the fault location, that is, using deep learning to analyze the data collected by the FTU, obtain the corresponding relationship between the fault location of the line and the data analysis results, and establish a model including the fault location of the distributed power distribution network. Combined with the Particle Swarm Optimization algorithm to quickly review the short-circuit location, and combined with the fault classification processing method to perform secondary refined fault positioning, screen non-fault areas, avoid complex physical models, and only need a small amount of FTU information. It is possible to pinpoint the location of the fault with great precision.

Images