CNN-DBN-based partial discharge fault diagnosis method

Abstract

The invention discloses a CNN-DBN-based partial discharge fault diagnosis method, which comprises the following steps: constructing a deep belief learning network based on a deep convolution neural network and a restricted Boltzmann machine model; collecting partial discharge simulation data and actually-measured partial discharge data, mixing a part of the partial discharge simulation data and actually-measured partial discharge data to serve as a training sample set, and mixing the rest partial discharge simulation data and actually-measured partial discharge data to serve as a test sample set; carrying out unsupervised training on the deep belief learning network by utilizing the training sample set, and extracting cross-mode features; inputting the cross-mode features into a logical regression classifier, and carrying out supervised training on the regression classifier by utilizing the test sample set to obtain a trained deep belief learning network; and inputting partial discharge to be tested into the trained deep belief learning network to obtain a partial discharge fault diagnosis result. Accuracy of fault diagnosis is improved.

Description

technical field [0001] The invention relates to a method for diagnosing partial discharge faults based on a CNN-DBN network, in particular to a method for diagnosing partial discharge faults of GIS combined electrical equipment based on a CNN-DBN network, and belongs to the field of device fault diagnosis. Background technique [0002] In the modern power system, the emergence of GIS combined electrical appliances provides a new solution for the construction of transmission and distribution stations. A GIS combined electrical appliance can package circuit breakers, isolating switches, voltage transformers, rheological transformers, lightning arresters, busbars, bushings, etc. A variety of electrical equipment, and the insulation distance is shortened by sulfur hexafluoride gas, saving space. Due to its high reliability, small footprint and long maintenance cycle, it is widely used in power transmission and transformation equipment of various voltage levels. [0003] However...

AI Technical Summary

Problems solved by technology

Existing methods must rely on certain data feature extraction, such as wavelet transform model and Hilbert-Huang energy transform model, GIS combined electrical partial discharge signal contains a large number of high-frequency components, and wavelet transform, Hilbert-Huang transform, etc. are sensitive to noise. certain sensitivity

Moreover, the features extracted by the method based on feature extraction do not necessarily reflect the actual defects

Images