Research On Autonomous Obstacle-Crossing Method And Control Strategy Of Inspection Robot Based On Multi-Sensor Information Fusion

The purpose of high-voltage line inspection robot is to provide a set of automatic operation system for the inspection of high-voltage transmission lines,instead of inspectors,so as to reduce the labor intensity,reduce the cost of inspection and improve the quality and efficiency of inspection.At present,the problems of low efficiency and poor stability in obstacle surmounting process commonly exist in existing inspection robots,which seriously restrict the popularization and application of inspection robots.With the development of control theory and the theory of multi-information fusion,it is one of the effective ways to improve the efficiency of autonomous inspection for robots to carry out the research on the optimization algorithm of autonomous obstacle surmounting based on multi-sensor fusion.In this paper,the theory of multi-sensor information fusion is applied to obstacle detection on the robot line and pose detection of the robot body.On-line obstacle detection,the laser radar,machine vision,infrared photoelectricity and other sensors are fused,the information matrix of obstacle types is established,and the method of line obstacle recognition and classification control is proposed.In the aspect of robot pose detection,the action that does not affect robot pose is simplified.The simplified kinematics model of the robot and the spatial description of obstacles are established.The position and posture of the robot are detected by position sensors on each joint,and the position and posture of the robot are detected by gyroscope.A position and posture detection method based on position and posture sensor information fusion is proposed to improve the accuracy of the position and posture detection.According to the distance information of obstacle type in front of the robot and the real-time position and pose information of the robot body,this paper studies an autonomous planning method of obstacle-crossing process based on fine-tuning obstacle-crossing sequence.The obstacle-crossing sequence of the robot is divided into several basic action units and the action correlation between adjacent action units is analyzed based on the action correlation.On this basis,a new obstacle-crossing process is planned and generated to improve the efficiency of obstacle-crossing.In order to verify the accuracy of obstacle detection and pose detection on line and the feasibility of fine-tuning obstacle-crossing sequence,a prototype of inspection robot has been developed.A series of experiments have been carried out,including multi-sensor detection module experiment,indoor and outdoor simulation environment debugging experiment and on-line inspection experiment of actual high-voltage transmission line robot.Compared with single sensor detection,the theory based on multi-sensor information fusion can improve the accuracy of On-line Obstacle Detection and ontology pose detection.Compared with the method before optimization of obstacle-crossing sequence,the method based on fine-tuning of obstacle-crossing sequence can improve the efficiency of obstacle-crossing and show good adaptability to the line working environment.