Research On Online Trajectory Generation,simulations And Experiments Of A Cobot

With the proposal of ‘made in China 2025',high-end CNC machine tools and robots have become one of the key areas of the manufacturing power strategy.Compared with traditional industrial robots,cooperative robots can adapt to the unstructured environment,so they have become the hotspot of robot research.In this paper,a 6-DOF ‘Bright' lab-developed cooperative robot is taken as the object,and the online trajectory generation method is mainly studied when the robot performs tasks in the unstructured environment so that the robot can change the current working state with the optimal velocity or acceleration peak.The research results can effectively improve the adaptability and response-ability of a cooperative robot to an unstructured environment.The main research contents and innovations are as follows:Firstly,based on the kinematic model of the 6-axis cooperative robot,the quaternion and Euler angle representation of the robot's attitude are analyzed.Both the calibration methods and the error analysis of the tool coordinate system and the workpiece coordinate system are considered.Then,the linear planning,the arc planning and the attitude interpolation of the tool center point of the robot are studied respectively.Secondly,for the problem that the cooperative robot can adjust the pose according to the external signal when working in an unknown environment,based on the study of the optimal online planning method for the minimum time of single-axis S-curve velocity profile,two time synchronization methods of multi-joints are proposed according to the application requirements of different occasions.For one thing,the velocity peak optimal multi-joint synchronous online planning method can change the current motion state immediately with the maximum capacity of the robot after receiving the new task and is suitable for the case where the end path has geometric constraints.For another,the time synchronization method with optimal acceleration peak is for certain transition areas without path constraints so that the joints of the robot can smoothly transition from the current motion state to the target motion state,and the motion trajectory is smoother than the first method.Finally,a simulation model of the 6-DOF cooperative robot is built on V-REP,and the online trajectory generation method is verified by remote API joint functions.Then Twin CAT FIFO communication mechanism is used to control the cooperative robot,both the two online planning algorithms proposed in this paper are verified by experiments.The experimental results show that the online planning methods proposed in this paper can quickly and effectively respond to any new target motion state.