Insulator string fault detection method based on infrared image temperature distribution law and BP neural network

Abstract

The invention provides an insulator string fault detection method based on infrared image temperature distribution law and a BP neural network. The method comprises the steps that the rectangular target area of an insulator string in an infrared image is extracted through manual and color space conversion; the length feature of the target area is calculated through image processing techniques such as binarization, Hough transform and rotation to extract insulator steel caps and disk surface areas respectively; an image temperature matrix and a segmentation result are used to calculate the average value of each steel cap and disk surface area as the feature quantity, and a K-Means clustering algorithm is introduced in the calculation process to remove background pixels; in the order of insulators from the low voltage end to the high voltage end, temperature eigenvectors of the steel caps and disk surfaces are formed respectively; and based on the temperature distribution law of the insulator string, an insulator string fault diagnosis model based on the BP neural network is established. The detection method provided by the invention has high recognition accuracy.

Description

technical field [0001] The invention relates to the technical field of detecting the operating state of power transmission and transformation equipment, and more specifically relates to an insulator string fault detection method based on infrared image temperature distribution law and BP neural network. Background technique [0002] As a special insulation control, the insulator string plays the role of mechanical support and electrical insulation, and has an important impact on the normal operation of substations and transmission lines. Insulator strings working under outdoor conditions are affected by natural environments such as wind, rain, and snow for a long time, and the insulation state ages, forming low-value and zero-value insulators; or affected by the combined effects of particles in the air and the natural environment, the surface accumulates pollution Cause flashover discharge; or cause insulator failure due to quality problems in the production process, affecti...

AI Technical Summary

Problems solved by technology

[0004] The traditional methods applied to the detection of low zero value insulators include short-circuit fork method, small ball discharge method, insulation resistance method, laser Doppler vibration method and voltage distribution method, etc. The traditional methods of pollution detection include measuring equivalent salt density or gray density , to measure the surface conductance of the insulator and to measure the leakage current. These methods have the advantage of high accuracy, but there are problems such as heavy workload, low safety, high cost, or incapable of live maintenance

Infrared detection technology is a non-contact passive measurement technology, which has the advantages of non-stop and fast speed. It has been widely used in the condition maintenance of various electrical equipment, such as transformer bushing heating and oil shortage fault detection, wire connection However, the problem existing in the current practical application of infrared detection technology is that it is difficult to accurately judge voltage-induced heating faults, mainly relying on the subjective experience of operation and maintenance personnel, reliable less accurate

At present, there are few studies on insulator fault diagnosis using this rule, and most of the research focuses on using infrared thermal imager analysis software to analyze the temperature curve of the central axis of the insulator string, and it is difficult to determine the fault type and fault location

Images