Deep-belief-network-based diagnosis method of distribution network

Abstract

The invention discloses a deep-belief-network-based diagnosis method of a distribution network. The method comprises: original monitoring data of a distribution network are obtained; denoising and normalized modeling are carried out on the original data; hyper parameters of a distribution network fault diagnosis model are set; 70% of collection data are used as training samples to train the model;the rest of 30% of samples are used for testing the trained fault diagnosis model, wherein six outputted nodes are used for expressing probabilities of occurrences of two kinds of faults at three phases respectively and training is carried out again if the precision does not meet the requirement; the operation state of the distribution network is monitored by the model; and if a fault is caused,a fault type and a line are obtained. According to the invention, the deep learning theory is applied to the fault diagnosis of the distribution network; and changing characteristics of all collectiondata under various faults are learned automatically under the circumstances of complicated structure, many devices, and frequent data deficiency. Meanwhile, the good fault tolerance performance is realized; the fault diagnosis accuracy and timeliness of the distribution network are improved; and thus the distribution network can work stably and safely.

Description

technical field [0001] The invention relates to a distribution network, in particular to a distribution network fault diagnosis method based on a deep belief network, which is used for on-line fault diagnosis of the distribution network. Background technique [0002] As the last link of the power system, the power distribution system is directly responsible for the requirements of users in terms of power stability, safety, quality, and economy. With the increasing development of our country's economic level, the improvement of our people's living standards and the application of a large number of precision household appliances, users have put forward higher requirements for the quality and reliability of power supply. [0003] When a fault occurs in the distribution network, a large amount of action information in the system is aggregated to the centralized control center, and the information will be distorted or even lost during transmission. In addition, it is impossible ...

AI Technical Summary

Problems solved by technology

[0005] (1) The expert system is to use the relay protection in the power grid, the action principle of the circuit breaker and the past fault finding experience of the dispatcher to form the knowledge base of the expert system in the fault diagnosis. According to the real-time alarm signal, the inference result is obtained according to the knowledge base, but This method is slow, has poor fault tolerance, and cannot learn independently;

[0006] (2) Artificial neural network, which uses relay and circuit breaker state information to calculate fault information, but this method has a large diagnostic error in a nonlinear power network;

[0007] (3) The power network fault diagnosis method based on Petri network and probability theory has the advantages of fast operation speed, high fault tolerance performance, and good accuracy, but the method is not effective for fault analysis with high timing requirements, and The state combination explosion of the model is prone to occur;

[0008] (4) The distribution network fault diagnosis based on Bayesian network combines prior information with posterior information, which can well reduce the subjective bias when only the former exists and the noise influence of only the latter, but it needs to find out the internal The conditional probability and prior probability of each event, and these data need to be obtained through tedious calculations

Images