Research On Trajectory Planning Method Of Measuring Manipulator For Complex Curved Workpiece

In recent years,3D vision measurement technology has developed rapidly.With the increase of measurement accuracy and efficiency,it has been widely used in various measurement fields and has become the mainstream trend of non-contact measurement in the field of reverse modeling.Aiming at the limitation of the working space and posture of the traditional linear structured light sensor and the single-degree-of-freedom displacement platform,the industry adopts the measurement method of the mechanical structure equipped with linear structured light,which solves the limitation of the pose and greatly improves the flexibility of the measurement system.However,due to the complex configuration of the measured object and the many surface features,manual teaching and measurement are required.It takes a long time and the measurement accuracy is difficult to guarantee.Therefore,it is particularly important to study a method that can intelligently plan the scan path.In response to the above problems,this paper uses the depth camera Kinect as the visual input of the scanning trajectory planning,and proposes a measurement method based on the depth camera perception.The Kinect is driven by another mechanical arm to perform a ring scan to obtain the rough surface point cloud of the measured object.Rough point cloud for scanning trajectory planning.For the measurement system,the measurement indexes of accuracy and speed are proposed,and in-depth theoretical research and experiments are carried out on the measurement system model,point cloud processing,surface reconstruction,path planning,etc.The main research contents are as follows:Firstly,the design of the measurement system is carried out,a complete mathematical model is constructed,the positive kinematics model of the robotic arm is established by the modified DH method,and the hand-eye calibration method to determine the radius ball by scanning determines the measurement coordinate system of the line structured light and the robotic arm The conversion relationship betweenthe terminal coordinate system discusses the line structured light and the imaging model of the general camera.Secondly,the research on point cloud data processing was carried out,and the work of point cloud filtering,segmentation,simplification,registration,surface reconstruction,normal vector calculation,and hole marking were realized.By reconstructing the surface with a rough point cloud,the scanning path planning problem is transformed into an approximate Hamiltonian path problem in the field of graph theory.The ant colony algorithm,the path search algorithm based on BFS,and the dijkstra path finding algorithm are used to obtain the scan path that meets the measurement index.Aiming at the rough point cloud surface void problem and the actual measurement angle difference,a dynamic scanning trajectory planning method is designed based on the point cloud boundary extraction.Finally,the model building and scanning simulation of the measurement system were carried out to verify the effectiveness of the measurement system,and the measurement accuracy reached 0.2mm.