User friendly Matlab-toolbox for symbolic robot dynamic modeling used for control design

摘要

In this paper a new Robot Modeling/Simulation Toolbox for Matlab is presented. The primary purpose of this toolbox is to generate all the common equations required for robot control design. It can compute the kinematic and dynamic equations of a serial robot in closed-form. The toolbox generates codes for the most representative matrices of the robot dynamics. For example, the Inertia Matrix, Coriolis Matrix, Gravitational Torques Vector and most important the Robot Regressor can be computed in closed-form with symbolic representation. This toolbox uses the Denavit-Hartenberg (DH) and Euler-Lagrange Methodologies to compute the Kinematic and Dynamic models of the robot. Furthermore, it automatically generates useful code for these models, such as M-Files, Simulink model and C/C++ code, allowing easy integration with other popular Matlab toolboxes or C/C++ environments. The only requirement from the user are the DH parameters, making it an easy to use tool. For 3D visualization, the toolbox supports different methods. The primary contribution is the automation and simplification of the robot modeling process which is important for correct robot design and control. In addition, the easy to use GUI and simplified models allow rapid prototyping and simulation of robots and control design/validation. As a proof of concept, validation of the computed models of a real industrial robot is included, where the toolbox was used to compute all the robot models. Thereafter, using the motion equations generated by this toolbox, a Dynamic Compensation Control was designed and implemented on a Staubli TX-90 industrial robot in order to demonstrate how this toolbox simplifies the process.