Open Partial Discharge Big Data Management System Based on Intelligent Diagnosis Algorithm

Abstract

The invention discloses an open partial discharge big data management system based on an intelligent diagnosis algorithm, which includes a data access module, a data conversion module, a database and a diagnosis analysis module. The data access module is used to provide a standard data input interface to receive Detection data; the data conversion module converts the data received by the data access module and normalizes it into the data format required by the platform; the database includes a data hierarchical storage module, and the data hierarchical storage module includes basic databases and advanced databases; diagnostic analysis The module is connected with the database, and is used for judging the discharge diagnosis result of the data to be diagnosed and using the advanced data set in the advanced database to train the intelligent diagnosis algorithm. The invention trains and optimizes the intelligent diagnosis algorithm, and provides data support for the diagnosis; the open data management enables the user to put the data and the diagnosis algorithm into the platform for verification; it has multiple diagnosis mechanisms, ensuring the reliability and accuracy of the diagnosis results sex.

Description

technical field [0001] The invention belongs to the field of data management, and in particular relates to an open partial discharge big data management system based on an intelligent diagnosis algorithm. Background technique [0002] With the continuous expansion and development of the national grid scale and the comprehensive development of the national smart grid construction, the requirements for maintaining the safe and stable operation of various power equipment in the power system have also been further increased. Among them, the method of partial discharge live detection is used to analyze and evaluate the insulation status of high-voltage power equipment, and the idea of ​​formulating corresponding maintenance strategies based on the analysis and evaluation results has been widely used. [0003] With the development of partial discharge live detection work, especially the large-scale development of paid service detection in recent years, the amount of detected parti...

AI Technical Summary

Problems solved by technology

One of the most common methods is to establish a data management system to manage data and transform from traditional paper management to digital management. However, many years of actual operation have found that there are still many problems in this management: 1. Due to partial discharge detection Different types of power equipment objects lead to various data files for partial discharge detection, including data files, picture files, and data recording tables; 2. The data itself is of various types, including pulse sequence phase distribution diagram PRPS / partial discharge phase distribution diagram PRPD data, the four-parameter data of contact ultrasonic and the ground wave data of switchgear, etc., and with the continuous improvement and development of detection technology, new data are constantly being generated, such as non-contact ultrasonic recording files; 3. For local There are many manufacturers of discharge detection equipment, and there are many types of products, resulting in inconsistent and incompatible data interfaces and data protocol formats of the equipment, and the detection values ​​are not properly normalized and calibrated

[0004] The above problems lead to the complex management of partial discharge detection data, which is difficult to manage and use, often leading to the shelving of data management or the lack of operation and maintenance, so that the "intangible assets" of partial discharge data have not only failed to "grow" with the development of information mining technology. Rapid value-added” is accompanied by the risk of “loss”

[0005] In terms of data analysis and application, some problems have also been found in many years of application: 1. The reliability and accuracy of data analysis are not high, especially in the differences in data analysis results, and the analysis results of the same test data from different manufacturers 2. Users cannot put their own diagnostic methods or data into the data management system for detection and verification; 3. The current judgment standard for partial discharge signal characteristics is too Single and fixed, such as threshold comparison and spectrogram, are still insufficient for the information mining hidden in the data; 4. The collection and processing of partial discharge data has insufficient accuracy or information loss due to hardware reasons such as collection rate; 5. Partial discharge data In the analysis and research, there is a lack of data with high validity, good reliability, strong correlation, recyclability and supplementation, and signal integrity and traceability, which leads to the inability of lasting, effective and cyclic iterative optimization in partial discharge analysis and research

Images