Online inspection identification method and system based on multi-source data

Abstract

The invention discloses an online inspection identification method based on multi-source data. The method comprises the following steps: training an identification model by using visible light pictures of primary equipment in different weather scenes; performing health recognition on the primary equipment visible light picture collected on site through the trained recognition model; performing health identification on the primary equipment through the collected infrared image of the primary equipment; performing health recognition on the primary equipment through the collected voiceprint of the primary equipment; performing health identification on primary equipment through sensing layer sensor data; performing health identification on the primary equipment by collecting primary equipment protection measurement data; judging the health condition of the primary equipment according to the health recognition result of the collected visible light pictures of the primary equipment in different weather scenes; the health condition of the primary equipment is judged according to the health recognition result of the visible light picture of the primary equipment in combination with the data of various data sources. The method can improve the judgment accuracy of the health abnormality of the primary equipment.

Description

technical field [0001] The invention belongs to the technical field of on-line monitoring of electric power equipment, and in particular relates to an on-line inspection identification method and system based on multi-source data. Background technique [0002] Online inspection is an important technical means to support the operation and maintenance of the transportation inspection business and the inspection work. The reasons include the following: [0003] 1) Suitable for the development needs of unattended substations [0004] With the continuous improvement of the production automation level of the power system, unattended substations have emerged at the historic moment. At present, many substations have added remote viewing functions on the basis of the functions of telemetry, remote signaling, remote control, and remote adjustment, so as to realize fire protection and personnel intrusion. , stray into restricted or dangerous areas, and introduce video images into the ...

AI Technical Summary

Problems solved by technology

[0011] To sum up, the existing substation monitoring image recognition methods all adopt the method of "complicated image preprocessing - artificial design features - calculation of artificial features - machine learning", but due to the problem of substation monitoring image training samples , modeling problems of weather and meteorological influences, feature extraction of different devices, etc., resulting in low recognition rate of some devices under certain weather conditions

Images