Study On Kinematics And Dynamics And Trajectory Planning Algorithm Of 4R Desktop Robot

With the rapid development of robot technology and the urgent needs of the market,China has listed the robot industry as a key research field of "Made in China 2025".This paper based on the Industry-University-Research Collaboration,the Dobot robot as an example,the developing of function expansion and redeveloped,and the research and development a teaching type 4R desktop robot,it is aimed at enable students to better understand and cognize the robot.In the research and development process,this paper focuses on the research of the kinematics,dynamics and trajectory planning algorithm of the desktop robot.In order to avoid the possibility of solution and ensure the accuracy of angle.According to the structure characteristics of the desktop robot pose,the robot joints coordinates and kinematics equation by D-H method,the kinematics analysis,obtained the inverse solution,and the dual variables arctangent function is applied to the expression of the solution the counter,the correctness of the forward and inverse solution is verified by real data.At the same time,analyze the differential motion of the robot,obtained the Jacobi matrix of the robot by the construction method.Kinematics analysis lays a solid foundation for the analysis and research of the dynamics and trajectory planning of desktop robot.According to the dynamic modeling method of traditional robot is the complex process,model solving difficulties,the big calculation error,and the difficulty increase rapidly with the degrees of freedom increase of robot.Inspired by artificial intelligence,this paper puts forward new method,which can help the robot to learn dynamic modeling.Using the joint variables as input and the torque values of each joint as output,the three layer BP neural network model is constructed to solve the dynamic inverse problem of the desktop robot dynamic.In order to improve the convergence speed and stability of the model,the elastic gradient descent algorithm is introduced.By using MATLAB programming,the predicted value of the torque corresponding to each joint is compared with the calculated value.The prediction results show that the desktop robot torque mainly borne by 2,3 joints,other joints force is small,the BP neural network model can predict the torque value of each joint of desktop robot in any posture.The use of artificial intelligence to solve the problem of robot dynamics is of great reference value for the future research of robot dynamics.In order to find the optimal path,the solution of the inverse solution and the motion stability of the robot are discussed.First,the flow chart of the trajectory planning algorithm for the desktop robot is designed,and obtain the starting point and the end point by teaching methods in the Solidworks.Second,the joint angle is obtained by using the linear interpolation algorithm and the inverse solution formula in Cartesian space,and useing the dynamic programming algorithm selected a sequence of optimal solutions.Finally,for the discrete optimal solution,the cubic B-spline interpolation is used to process it.The simulation results show that the optimal motion path and the stability of the robot motion are obtained,and the correctness of the algorithm is verified by MATLAB.In order to improve the efficiency of robot design.This paper use MATLAB GUIDE to develop the robot analysis and simulation platform.The platform realizes the analysis and simulation function of the robot forward kinematics and inverse kinematics,the forward and inverse dynamics problem,the animation demonstration and the trajectory planning.And the simulation results are output in the form of numbers,and the visual effect is high.On the basis of kinematics analysis.The inverse dynamics problem is solved by using BP neural network.The optimal path is obtained by using linear interpolation,dynamic programming and cubic B-spline interpolation algorithm.The development of simulation platform,the design efficiency of desktop robot is improved effectively.The research results of this paper can provide a reference for the design and research of robot.