Power distribution network fault detection method based on convolutional neural network

Abstract

The invention provides a power distribution network fault detection method based on a convolutional neural network, which provides the technical support for the three problems of a fault type, a faultphase and a fault position, and comprises the following steps of S1, collecting a power distribution network fault data set; S2, performing data preprocessing on the current sequence data by utilizing the simplified Hilbert-Huang transform; S3, initializing a convolutional neural network designed by the patent; S4, training the convolutional neural network by using the preprocessed data; and S5,testing the fault detection method by using the test set. According to the method, by utilizing the convolutional neural network and the simplified Hilbert-Huang transform, the fault type, phase and position of the power distribution network can be accurately judged, and meanwhile, compared with a method for carrying out fault judgment by utilizing a recurrent neural network, the speed is greatlyincreased.

Description

technical field [0001] The invention belongs to the technical field of deep learning, and in particular relates to a distribution network fault detection method based on a convolutional neural network, which uses a convolutional neural network and a simplified Hilbert-Huang transform to detect distribution network faults. Background technique [0002] In today's society, electricity, as the main energy source, plays a vital role in daily work and life. With the development of urbanization, the area and equipment using electricity are also increasing rapidly. Therefore, it is necessary to build a reasonable and reliable distribution network. In 2016, the plan for power construction was proposed, emphasizing that smart grid should become the trend of future grid development, and the planning will also provide corresponding financial support for the construction of smart grid in towns and villages. It can be seen that the construction of smart grid has received the attention o...

AI Technical Summary

Problems solved by technology

Compared with other methods using decision trees or random forests, great progress has been made in accuracy, but it still cannot meet the requirements of fast and accurate fault judgment

[0006] At present, there is no description or report of the similar technology of the present invention, and no similar data at home and abroad have been collected yet.

Images