Research Of Visual Navigation Applied On High-voltage Inspection Robot

The high-voltage transmission line is the lifeline of modern industrial society and national life.Whether it is normal or not is related to national livelihood and social stability.However,the way of long-distance transmission and the special environment will inevitably cause line failures and cause great loss of life and property.Therefore,regular inspections of high-voltage power lines have become a necessary task for transmission departments.The development of the inspection robot provides a new solution for the inspection work of electric power.The major difficulty of the inspection robot is the design of its autonomous navigation system.This subj ect conducts research on the visual navigation system of the high-voltage inspection robot.The research focus is on the design of the visual navigation algorithm.It is expected that the high-voltage inspection robot will work autonomously and continuously along the line.First,in view of the problems of low detection accuracy,weak anti-interference ability,and poor detection effect of small targets in existing visual inspection algorithms for high-voltage transmission line inspection robots,a nearest neighbor weighted gray scale based on polynomial function curve fitting is proposed Mean algorithm,used to extract the target region of interest(Region Of Interest,ROI).Experimental results show that the algorithm has great ROI extraction accuracy.Secondly,in view of the problems of the traditional target detection algorithm,such as multiple manual participation and complex feature extraction,based on the YOLOv3 target detection framework,some network layers are deleted,and Focal Loss is used to improve the original loss function.Experimental results show that the algorithm greatly improves the recognition rate of positive samples on the basis of a significant increase in running speed.Thirdly,in order to solve the problems of the current target positioning algorithm,such as weak anti-interference ability,slightly lower positioning accuracy,too many a priori conditions,etc.,a stereo matching algorithm based on line matching is proposed,and through the distribution information of distance constraints and depth values,By removing some matching points,the weighted average depth of the target is obtained.Experimental results show that the positioning algorithm has great anti-interference ability and high positioning accuracy.Finally,in order to provide an embedded platform for the operation of the visual navigation algorithm,according to the requirements of the visual navigation system,the embedded structure of the visual navigation system of the high-voltage inspection robot is designed,and the software program is written to complete the established functions of the system.