A method and apparatus for real-time detection of infrared image abnormality of electric equipment in substation

Abstract

A method and apparatus for detecting infrared image abnormality of electric power equipment in substation The invention adopts the latest depth learning and the computer vision algorithm to automatically analyze the infrared data and detect the abnormal heat through the collected infrared image or infrared video, which will greatly improve the efficiency and accuracy of the infrared image or infrared video detection. In the substation patrol robot with infrared imager, such intelligent infrared abnormal heating area and fault equipment type detection algorithm based on depth learning are embedded in the upper computer control system such as substation fixed-point monitoring track robot. Real-time infrared image or video detection will further improve the intelligent degree of these intelligent equipment, so as to achieve the real intelligent patrol.

Description

technical field [0001] The detection technology field of the present invention specifically relates to a real-time detection method and device for infrared image abnormality of power equipment in a substation. Background technique [0002] Power equipment is the basic component module of the substation. Timely detection or real-time monitoring of the status of power equipment can efficiently manage these equipment, making the substation more efficient, safe, and reliable. Power outages, thereby greatly reducing the impact of power outages on social, economic, production and life. The abnormal heating of most of the power equipment in the substation is due to poor contact of key parts, loose connection parts, overload, oxidation of old equipment joints, leakage current and other reasons, which cause abnormal heating in some areas of power equipment. Infrared thermal imager detection has the advantages of non-contact and non-stop power detection of electrical equipment, and i...

AI Technical Summary

Problems solved by technology

Infrared thermal imager detection has the advantages of non-contact and non-stop power detection of electrical equipment, and is widely used in various power units. With the application and popularization of intelligent equipment such as substation intelligent inspection and fixed-point monitoring, infrared image data continues to increase, and massive inspections Infrared image data is retrieved and analyzed manually, which will be time-consuming and labor-intensive and inefficient

The method of scanning infrared image pixels to find the highest temperature point has defects such as large temperature changes, difficult adjustment of temperature thresholds, large amount of stored data, and inability to sensitively monitor and identify faulty equipment; traditional image processing feature analysis algorithms such as SIFT, SURF, ORB, HOG, etc. are difficult to extract and identify the image features of irregularly shaped heating areas in infrared images

[0003] In the past, infrared detection methods often required manual setting of infrared temperature thresholds or manual observation to judge abnormal fever, which was time-consuming and laborious, and could not be continuously monitored.

The method of using traditional image processing technology to identify abnormal heating of power equipment is often a single task, and the accuracy of identification and practicability are also a big problem

Images