Research On Motion Control Of Bridge Detection Flying Wall-climbing Robot

This article combines the graduate joint training base project(JDLHPYJD2021007)collaborated with Chongqing Intelligent Robotics Research Institute to design and develop a detection robot that combines flight and wall climbing for the needs of economic development and complex bridge environments.Analyzed the posture transformation of the robot under different working conditions,established a dynamic model of the robot,designed a robot trajectory tracking control system,designed a robot overall design scheme that meets the design criteria,and designed a fusion filtering algorithm for experimental research on bridge crack detection.Firstly,a literature review was conducted on the current development status of bridge detection and wall climbing robots both domestically and internationally,and the problems and challenges faced by wall climbing robots in bridge detection at the current stage were analyzed.Based on the structural and kinematic characteristics of the flying wall climbing robot,the attitude transformation process of the robot was analyzed,and a dynamic model of the robot including a moving mechanism and an adsorption mechanism was established.To improve the work efficiency of flying wall climbing robots,a robot trajectory tracking control system has been designed,which enables the robot to stably track the preset trajectory under safe adsorption conditions.The robot coordinates,speed,heading angle,and adsorption force are obtained through sensors in the robot perception layer,and a preset reference trajectory is added as input to the controller.The rotor speeds of the driving wheels on both sides and the adsorption device are used as output to control the robot to travel along the preset trajectory.According to the deviation between the actual coordinates,speed,heading angle of the robot and the preset trajectory,the input of the motion mechanism is calculated through a fuzzy controller.To verify the effectiveness of the designed controller,MATLAB/Simulink simulation analysis was conducted on the linear and circular preset trajectories.The results indicate that the trajectory tracking control system designed in this thesis can quickly adjust the adsorption force feedback and accurately track the trajectory.Design the overall structure and motion characteristics of the flying wall climbing robot,including the main control module,adsorption module,motion module,and environmental perception module.Select the hardware of each module and design corresponding circuits and software control schemes.A bridge crack image filtering method based on adaptive weighted mean filtering algorithm is proposed for bridge images filled with complex noise.This method combines the characteristics of grayscale stretching and adaptive weighted mean filtering,utilizing grayscale stretching to balance image shadows and lighting,and using an improved mean filtering method to remove complex noise.At the same time,the Otsu method and binarization technology based on connected domain filtering are used to remove isolated noise points,and quantitative analysis of crack area,length,and width is carried out by combining the crack skeleton map.The experimental results show that the designed fusion filtering algorithm can effectively remove noise and improve the accuracy of crack detection.In summary,this article establishes a dynamic model of a flying wall climbing robot,studies the trajectory tracking control method of the flying wall climbing robot,designs the overall scheme of the robot,and studies the fusion filtering algorithm based on the complex characteristics of bridge images.The research work of the thesis provides reference for the scheme design,motion control research,and image detection system design of robots.