Research On Trajectory Planning And Positioning Method Of Mobile Sandblasting Robot

Sand blasting is an indispensable means in the anti-corrosion treatment of large-scale steel structure workpieces.At present,almost all sand blasting operations for such workpieces in China are manual operations,with many problems such as backward production mode,high labor intensity of operators,high incidence of occupational diseases such as silicosis,etc.Therefore,the development of intelligent automatic sandblasting system has become an urgent demand of related industries.In this paper,the key technologies of the automatic sandblasting system for large-scale steel structure workpiece are studied,including the research of workpiece positioning system based on binocular vision,the research of image processing in sandblasting environment,and the research of trajectory planning of robot arm of sandblasting robot.Firstly,the research of workpiece positioning system based on binocular vision is completed.The image acquisition method of the binocular vision positioning system is designed,the camera of the binocular vision positioning system is calibrated,the internal and external parameters of the camera are determined,the feature extraction method of the workpiece and the accuracy optimization method are determined.Through the Halcon platform,the feature points of sandblasting workpiece are matched,the three-dimensional reconstruction of binocular vision positioning system is realized,and the workspace model of sandblasting robot is established by transforming it into the coordinate system of sandblasting robot.Secondly,aiming at the problem of image processing in sandblasting environment,an image processing method is proposed.Through the analysis and modeling of the causes of image quality degradation in sandblasting environment,the image attenuation model in sandblasting environment is obtained.Based on the analysis of the applicability of the existing image processing methods in the sandblasting environment,compared with other environments,the brightness of the processed image in the sandblasting environment is too low.Aiming at this problem,a processing method to improve the brightness of the image in the sandblasting environment is proposed.The results show that the method can get better visual effect after processing the degraded image in sandblasting environment.Finally,the trajectory planning of the robot arm is studied.A path planning method is formed,which is based on the passing point of the blasting gun and aims at the blasting quality.According to the working requirements of the sandblasting robot under different working conditions,the general path planning method is improved,the sandblasting process requirements are introduced into the path planning,the path point planning model for the sandblasting robot is established,and the traverse path planning for the sandblasting workpiece by the sandblasting gun is realized.The trajectory planning of sandblasting robot is based on path planning,and linear interpolation is used to ensure the quality and efficiency of sandblasting.This paper studies the collision detection of sandblasting robot in the working space,and puts forward the judgment conditions of collision detection of sandblasting robot.The trajectory of sandblasting robot is simulated by experiments.