Intelligent diagnosis method for local overheating of internal conductor of GIS basin-type insulator

Abstract

The invention relates to an intelligent diagnosis method for the local overheating of an internal conductor of a GIS basin-type insulator, and the method comprises the steps: 1, building a photographing posture calibration model based on standard visual image marking; 2, building a visual indexing authentication structure for a GIS switch basin-type insulator: designing a GIS switch basin-type insulator indexing mechanism based on geographic space meshing through combining with the characteristics of a power grid system, carrying out the coarse mesh dividing of a to-be-indexed target accordingto a geographic space, building a first-level index based on a fixed mesh so as to reduce the number of cross-mesh space objects, carrying out the fine dividing of the space objects completely located in a mesh region, building an index list for the cross-mesh space objects, building a high-efficiency database for the basin-type insulator indexing authentication, and embedding the position information codes into a visual indexing authentication model of an image mark; 3, carrying out the recognition of an internal abnormality.

Description

technical field [0001] The invention belongs to the field of image recognition, and relates to a comprehensive utilization of traditional image processing methods and deep learning object detection algorithms, based on multi-modal image data, for intelligent retrieval of GIS shell pot insulators, and for local hot spots of internal conductors. method of automatic diagnosis. Background technique [0002] Substations play a very important role in the power system, where voltage and current are exchanged and distributed, and then delivered to consumers. It is directly related to the electricity consumption of millions of households, and large-scale power outages will bring immeasurable losses to the national economy. With the steady development of my country's power system towards UHV, large capacity, and intelligence, the requirements for safe and stable operation of power equipment continue to increase. As an important equipment for collecting, distributing and transmitting...

AI Technical Summary

Problems solved by technology

As a kind of object detection task, the pot insulator detection task based on visible light image and infrared image depends on the proper selection and extraction of multi-modal image features. The existing single machine learning algorithm, template matching, feature matching and other methods cannot detect effectively

Images