Intelligent fault classification and location method for ultra-high voltage direct current transmission line

Abstract

The invention discloses an intelligent fault classification and location method for an ultra-high voltage direct current transmission line, and belongs to the technical field of relay protection of power systems. The method comprises the following steps of: classifying fault data by using a neural network by adopting a layered and distributed neural network model; distinguishing fault types; sending the classified data into different neural networks respectively for performing fault location; when the direct current transmission line has a fault and a sampling frequency is 10 kHz, selecting a discrete line mode voltage signal which has the sampling sequence length of 100 after the fault and performing S-transform, wherein a transform result is a complex time-frequency matrix of 51*100; solving the modulus of each element in the complex matrix to obtain transient energy distribution of the line mode voltage at all frequencies; selecting first five energy spectrums as sample properties; selecting a transfer function and a learning rule; setting proper neural network parameters for constructing a BP network model; and performing fault classification and fault location. A large number of simulation results show that the method has a good effect.

Description

technical field [0001] The invention relates to the technical field of electric power system relay protection, in particular to an intelligent fault classification and distance measurement method for UHV DC transmission lines using S-transformed transient energy. Background technique [0002] At present, there is no research on intelligent fault diagnosis of UHV DC transmission system in China, and there are very few intelligent fault diagnosis systems for UHV DC transmission system abroad, and there is still a certain distance from practicality. Most of the existing DC line distance measurement systems use the traveling wave method using the traveling wave propagation characteristics of the transmission line, that is, the distance to the fault is calculated by measuring the propagation time of the traveling wave on the transmission line, which is divided into single-ended method and double-ended method. Although the traveling wave ranging method has the advantage of not bei...

AI Technical Summary

Problems solved by technology

[0003] (1) The single-ended method has a small investment, but the subsequent reflected waves are easily affected by interference signals and difficult to identify; the double-ended ranging method needs to install ranging devices at both ends of the line, and requires time synchronization (GPS clock) and communication channels , the device is complicated and the investment is large

[0004] (2) The traveling wave signal generated by the fault has a large random component, is susceptible to interference, is fleeting, non-repeatable, and difficult to accurately measure and capture

[0005] (3) The traveling wave is affected by the smoothing reactor and DC filter at the end of the DC line, and there is a dead zone in the near area

[0006] Therefore, the DC line distance measurement based on the transmission line traveling wave propagation characteristics is easily affected by the above factors, and the distance measurement accuracy is poor

Images