Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Classifier creating method and partial discharge failure mode identifying method for transformer

A classifier and category technology, applied in character and pattern recognition, instruments, measuring electronics, etc., can solve the problems of pattern recognition accuracy and time, a large amount of training data and training time, etc., to shorten the recognition time and ensure The effect of accuracy

Inactive Publication Date: 2017-12-26
STATE GRID SICHUAN ELECTRIC POWER CORP ELECTRIC POWER RES INST +1
View PDF5 Cites 9 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

These algorithms can achieve relatively good results, but it is worth noting that the pattern recognition of partial discharge is often difficult to balance both accuracy and time.
To achieve better training results often requires a large amount of training data and training time

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Classifier creating method and partial discharge failure mode identifying method for transformer
  • Classifier creating method and partial discharge failure mode identifying method for transformer

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0028] A classifier creation method, including the following steps:

[0029] Multiple sets of characteristic data of three types of partial discharges, namely corona discharge, air-gap discharge and creeping discharge, are respectively selected as the training set and test set. Each characteristic data includes N attributes, where N is a natural number greater than 1;

[0030] The training set is resampled using the Bootstrap method, and the training set is randomly generated;

[0031] Use each training set to generate a corresponding decision tree. Before selecting attributes on each non-leaf node, m attributes are randomly selected from N attributes as the split attribute set of the current node, and the classification accuracy of these m attributes is The highest split method splits the node, where m is less than N; the split method is to select m attributes from the N attributes to classify the training set, where the value of m and which m attributes are selected The com...

Embodiment 2

[0049] A transformer partial discharge fault mode recognition method includes the following steps: input feature data into the classifier created by the method of embodiment 1, so as to output recognition categories.

[0050] When the classifier created by the method in Example 1 was used to identify the transformer partial discharge fault pattern, its input neurons were 24, representing 24 attributes of the partial discharge spectrogram respectively, and the output neuron number was 3, They are corona discharge, creeping discharge and air gap discharge.

[0051] In order to verify the accuracy of this scheme, 300 sets of characteristic data of known discharge failure modes are input into the classifier. The classification performance of random forests was analyzed using a three-dimensional graphical approach. According to the above steps, when using random forest to train the training set, a total of 1500 decision trees are generated, and there are only three output categori...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a classifier creating method and a partial discharge failure mode identifying method for a transformer. The classifier creating method includes steps of selecting multiple groups of feature data of partial discharge of three types including corona discharge, air gap discharge and creeping discharge as a training set and a testing set, wherein each feature data includes N attributes and N is an integer greater than 1; re-sampling the training set and generating training sets randomly; utilizing each training set to generate corresponding decision trees, picking m attributes randomly from N attributes as a split attribute set of a current node and splitting the node by adopting a splitting method with the highest classification accuracy among the m attributes before selecting attributes from each non leaf node; utilizing the decision trees for testing the testing set and obtaining corresponding types, and taking the type of the most output in the decision types as the type of the testing set. By utilizing a classifier created by adopting the method provided by the invention for identifying partial discharge failure modes of the transformer, accuracy is ensured and time is shortened.

Description

technical field [0001] The invention relates to the field of state diagnosis of power equipment, in particular to a method for creating a classifier based on a random forest algorithm and a method for recognizing a partial discharge fault mode of a transformer. Background technique [0002] Transformer partial discharge is one of the main causes of transformer insulation aging, damage, and electrical accidents. Therefore, it is of great significance to improve the reliability and safety of power transformer operation to improve the diagnosis method of transformer partial discharge in operation. [0003] At present, the fault diagnosis of transformer partial discharge is mainly based on threshold diagnosis, that is, when the discharge exceeds a certain set minimum discharge warning value, the system will give an early warning, and then the operation test personnel will judge and post-process. However, the threshold method is relatively single, and the amount of information t...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/62G01R31/12
CPCG01R31/1227G06F18/2148G06F18/24323
Inventor 冯运陈凌甘德刚张宗喜蒋伟
Owner STATE GRID SICHUAN ELECTRIC POWER CORP ELECTRIC POWER RES INST
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com