Dynamic Optimization And Application Of Robot Motion Trajectory Based On Binocular Vision

With the wide application of robots in digital flexible production lines,the ability to perceive the external environment and to dynamically optimize the trajectory of robots according to information is the perceive and key problem for the robot to realize autonomous motion.Although many scholars have explored the dynamic optimization of robot motion trajectory from various angles and obtained some results,there are still many problems to be further studied.In this dissertation,under the background that machine vision technology is becoming more and more perfect and popular in various fields,we combine binocular vision technology and path planning technology to the dynamic optimization of motion trajectory.This dissertation completes the following contents.(1)Firstly,the research background and significance of this topic are clarified.The research status of industrial robots,machine vision technology,path planning technology and motion tracking technology are studied.(2)Secondly,the model of the camera,the parallax principle of binocular camera and the limit restriction are introduced.The spatial cloud information of the scene area is obtained by semi-global stereo matching algorithm.According to the mapping relationship between camera coordinate system and world coordinate system,the point cloud is projected into robot coordinate system to complete scene reconstruction.(3)Thirdly,for the dynamic optimization of robot trajectory,this question can be divided into two technical points:1.path planning;2.motion detection and tracking.In the research of path planning technology,on the basis of the traditional A-star algorithm,the redundant nodes and inflexions in the path are eliminated,and the path is smoothed by B-spline curve to achieve the optimal and smooth trajectory of the robot.In the research of motion detection and tracking,the improved Mean-shift algorithm is used to real-time monitor and track the obstacles that may exist in the running process of the robot.According to the movement of obstacles,the path is discussed separately to realize obstacle-free optimal trajectory dynamic output and path searching of robot.(4)Then,the robot system is introduced.Through the analysis of the system requirement of this subject,modular design of the hardware platform of the robot based on binocular vision is carried out.And the connection mode,function and parameter selection of each module are introduced in detail.Thus the framework of the overall scheme of this subject is determined.(5)Finally,on the self-built hardware platform,two software platforms of binocular vision perception and motion control are designed.PC is used as upper monitor.Communication and real-time data interaction are completed through the self-compiled upper computer interface and motion control program.Thus the validity of the dynamic optimization of robot motion trajectory based on binocular vision method is verified.