Visual Simulation for Inertial Parameter Estimation of Robots by Base Sensor Measurements

Aiming to a problem of low identification accuracy because of the coupling of inertial parameters and joint friction parameters, a parameter identification schemes by the base force sensors measurement is proposed. Using the Newton-Euler dynamic iterative equations, linearization and solving with minimal parameter sets, Algorithm of robot inertia parameter identification model by force sensor measurements is derived and corresponding identification steps is given. In order to further test the estimation algorithm, a visual simulation of robot parameter estimation based on Simscape is studied. According to the actual robot system, the simulation modules of robot control, data acquisition, multi-body physics and visual display are designed in detail. Finally, actual control parameters and excitation trajectories are applied to the simulation system and the entire identification simulation is executed. The results show that the identification simulation system is correct and the inertial parameter obtained are accurate.