Research On Installation Method Of Cable Based On Vision

Using robots to assemble flexible parts such as cables is the key to further improving production efficiency and reducing labor costs.Many industrial fields have realized automatic assembly of workpieces with robots.Because flexible workpieces such as cables,flexible circuit boards,and seals are prone to deformation when subjected to external forces,the robot cannot predict and estimate the deformation of the workpiece or use the deformation of the workpiece to complete the assembly task.The assembly of deformable workpieces cannot be performed smoothly only by robots,and still needs to rely on the operation of skilled workers to complete.Aiming at these problems,this subject has carried out research on cable shape control technology,cable obstacle avoidance shape planning,and cable automatic assembly methods.Aiming at the problems that may occur during the cable assembly process,relevant solutions are proposed.In order to obtain the position of the cable in space,a vision-based cable geometry measurement method is used.The points on the centerline of the cable are sampled uniformly from the color image and registered with the depth map to generate a point cloud of the cable.The cable point cloud is then fitted using a Bezier curve to calculate the cable’s geometric shape.Aiming at the possible interference between the cable and the obstacle during the installation process,a vision-free cable shape controller was designed using a model-free linear soft body shape control method.In the shape control system,the error of the control point of the cable curve is the error between the desired control point and the actual control point.Combining the deformed Jacobian matrix and the robot Jacobian matrix,the motion control amount of the robot is calculated,and the cable shape control is realized.In order to achieve automatic obstacle avoidance of cables,it is necessary to automatically plan the target cable shape to avoid obstacles.Two methods are used to modify the Bezier curve model and the fixed-length curve model,and the bending deformation and torsional deformation of the curve are included in the optimization target,and the target shape of the cable is planned.A robot controller based on artificial potential field method is designed to guide the robot to avoid obstacles and complete the installation of cables.In order to verify the effectiveness of the above method,a visual servoing experimental platform based on the robot operating system(ROS)is set up in this paper.Vision-basedunmarked cable shape control experiments and artificial potential field method-based automatic cable installation experiments were performed on the platform.The effectiveness of the cable obstacle avoidance shape planning algorithm was verified,and the experimental results were analyzed.