Research On Theory And Application Of Robotic Grinding Accuracy

The definition of the robotic grinding—— the robot is used as the moving platform of tool(or workpiece),and the grinding tools with high-speed rotation are used to machine the surface of the part with controllable parameters.In actual processing,the robotic grinding systems are mostly composed of equipment such as the industrial robot,the workbench and the grinding tool.Among them,industrial robots mostly adopt a series of six-degree-of-freedom structures with poor rigidity;the contact wheels in grinding tools mostly use elastic rubber.Therefore,the robotic grinding system can be defined as a kind of non-rigid machining.The processing effect is affected by many factors,and one of the most important factors is the robotic grinding accuracy.In order to improve the robotic grinding accuracy,this paper focuses on the analysis of robotic grinding system,robotic kinematic parameter identification and error compensation,path planning method of complex curved surface parts,adaptive impedance control of grinding force and robotic grinding test.The main research contents are as follows:(1)A typical robot grinding system is designed and built by using the industrial robot,the force sensor,the grinding tool and the workbench.Based on this,the forward kinematics,inverse kinematics,and reachable working space of the industrial robot are analyzed using DH method,numerical method based on Jacobian matrix and Newton iteration,and Monte-Carlo method,respectively.Then,using the fast calibration method of tool center point(TCP),the representation method of target position and orientation,and the extraction and application of data points,a robot grinding method based on 3D model is proposed.Finally,the main influencing factors of robotic grinding accuracy are analyzed,which lays the theoretical basis and experimental foundation for the follow-up study.(2)To decrease the movement uncertainty of the TCP of the grinding robot,a method based on genetic tabu search algorithm is proposed to identify and compensate the structural parameters of the robot.Firstly,the error model and the unknown numbers to be calculated are determined by using the kinematic model and Jacobian matrix of the grinding robot.Secondly,a global search was performed using test data and genetic algorithms to obtain a global optimal solution.Then,the optimal solution is used as the initial solution of tabu search algorithm for further local search to obtain the final solution.At the same time,to accelerate the convergence speed,corresponding fitness function,dynamic crossover rate,dynamic mutation rate and dynamic search radius are set.Finally,a parameter identification algorithm based on the least square method is selected for contrast experiment.The results show that this method can effectively identify all kinematic parameters of the grinding robot and improve the absolute position accuracy of TCP.(3)To improve the surface quality,a novel path planning method is proposed in this article,which is based on the refinement algorithm.Firstly,the theoretical grinding trajectory is obtained by selecting the key contact points on the 3D model and calculating the inverse solutions of the B-spline curves.Next,based on the curvature variation rate and curve length criterion,the key contact points are further refined to produce the target points in the sensitive area of the workpiece where surface curvature changes abruptly or surfaces intersect.Finally,the grinding orientation at each target point location is obtained according to the solutions of the surface equation constructed using bicubic B-spline interpolation.To validate the method,a faucet is used for contrast experiment.The results show that this method is better than drag teaching method in planning efficiency and surface quality,and can reduce the surface roughness value of the processed workpiece.(4)In order to optimize the contact quality between the grinding tool and part surface,an adaptive impedance control of grinding force based on position is proposed.Firstly,the relationship between the normal grinding force and the tangential grinding force is established by the theoretical model of grinding force of robotic grinding.At the same time,according to the force analysis and integration scheme of the force sensor,the gravity compensation method of the grinding tool and the filtering method of the force signal are determined.Then,according to the feasibility of the second development of the grinding robot,the impedance control strategy based on position is established.During this period,in order to improve the machining quality of the system,the corresponding adaptive impedance controller is designed.Finally,the robotic grinding experiment is carried out with 3D printing rings.The results show that this method can effectively reduce the control error of grinding force and the flatness of parts surface.(5)In order to verify the overall performance and comprehensive application effect of the above methods,a robotic grinding experiment is designed for the post-processing of 3D printing blades.By comparing with the conventional grinding method,it is shown that the robotic grinding precision improvement method can improve the parameters such as the machining allowance,the overall shape,the surface roughness and the grinding force distribution of the complex surface parts,which can effectively improve the robotic grinding accuracy.