Power transmission tower bird-related fault state identification and evaluation method based on Mask RCNN

Abstract

The invention discloses a power transmission tower bird-related fault state identification and evaluation method based on Mask RCNN. The method comprises the following steps: 1, establishing an image database; 2, building a MaskRCNN power transmission tower fault state recognition and evaluation model, modifying the classification number of the last FC layer of the model, and carrying out the migration fine tuning of the model through an image database, and achieving the multi-target recognition and instance segmentation of the faults of birds, nests, insulators and bodies of towers; in combination with an image processing technology, correcting results of recognized bird droppings and bird pecking traces, and evaluating a fault state; 3, analyzing bird-related fault features, including bird-related fault type analysis and pernicious bird species feature analysis; and 4, establishing a bird-related fault discrimination mode fused with pernicious bird species discrimination by using an analysis result in the step 3, and realizing identification and evaluation of all types of bird-related faults. According to the method, safe operation of the power transmission line can be protected in a more refined manner.

Description

technical field [0001] The invention belongs to the technical field of image processing, and relates to a method for identifying and evaluating bird-related fault states of power transmission towers based on Mask RCNN. Background technique [0002] In recent years, the number of bird reproduction has increased sharply and the population has increased. The growing bird population has brought challenges to the safe and stable operation of transmission towers and their lines. According to statistics, during 2004-2006, the number of trips due to bird damage on 220kV and above transmission lines accounted for about 1 / 10 of the total trips, of which 220kV lines were the main ones; during 2010-2014, 27 provincial-level According to the survey results of the State Grid Corporation of China, transmission lines with a voltage level of 110kV and above have tripped 1,702 times due to bird-related faults; at the same time, the above data have shown an increasing trend year by year, and b...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to provide a method for identifying and evaluating the state of bird-related faults in transmission towers based on Mask RCNN, which solves the problems of poor robustness and weak generalization ability of image processing algorithms in the prior art; The problem of the early warning closed-loop mechanism has not been formed yet

Images