Infrared image-based substation equipment defect intelligent diagnosis method and system

Abstract

The invention discloses an infrared image-based intelligent substation equipment defect diagnosis method and system, and the method comprises: employing a YOLO deep learning model and a data analysismethod to achieve the automatic recognition and intelligent defect diagnosis of a substation equipment part in the infrared image through the steps of substation equipment part positioning, temperature feature extraction and temperature threshold judgment. On one hand, depending on the target detection performance of the YOLO deep learning network, the accuracy of substation equipment part positioning is ensured; on the basis of accurate positioning of power transformation equipment components, characteristic parameters are extracted, and the infrared image data is converted into a temperaturecharacteristic value capable of being quantitatively analyzed; on the other hand, depending on detection data and through the extracted temperature characteristic values, the distribution characteristics reflecting massive temperature difference data are obtained and analyzed, so that the defect threshold judgment method with higher reliability is obtained, the problems of a diagnosis method of pure image analysis or manual threshold setting are solved, and the reliability and the accuracy of infrared image defect diagnosis of power transformation equipment are improved.

Description

Technical field: [0001] The invention belongs to the technical field of fault diagnosis of electric substation equipment, and in particular relates to an intelligent diagnosis method and system for faults of electric substation equipment based on infrared images. Background technique: [0002] Infrared detection is a common detection method for substation equipment, and the defect diagnosis of infrared images is an important means for real-time control of the operation status of substation equipment. With the continuous expansion of the scale of substations in recent years, infrared detection is a routine detection method, and Shandong Electric Power Company alone produces millions of pieces of detection data a year. Therefore, relying solely on manual detection is obviously unable to meet the purpose of real-time control and analysis of the status of substation equipment. [0003] At present, the infrared diagnosis method for substation equipment is mainly based on the tra...

AI Technical Summary

Problems solved by technology

There are two main problems in this scheme. 1. ostu and region growing are traditional image processing methods, which have high requirements for the background of the substation equipment in the infrared image; 2. It is greatly affected by the background environment, and the segmentation accuracy is difficult to guarantee

In this case, it will seriously affect the accuracy of temperature extraction and defect diagnosis based on the results of traditional equipment inspection methods

[0005] Due to the influence of actual detection conditions, the traditional diagnostic method based on the DLT664 guidelines is relatively broad, without considering the influence of the detection environment, time, voltage level, etc., making the reliability of the diagnostic threshold low.

Images