An Online One-Step Method to Identify Inertial Parameters of the Base and the Target Simultaneously for Space Robots in Postcapture

摘要

Space robots are in free-flying or free-floating mode, motions especially attitude motion of the base and motion of the manipulator are strongly coupled. Regarding to the uncertainties of the target's inertial parameters and variation of the base's inertial parameters, this paper presents a novel online one-step parameter identification method to estimate all the inertial parameters of the target and the base simultaneously. Momentum- and force-based identification equations are derived from the linear and angular momentum equations of the system and their derivation, and the modified recursive least square method is used for solving the equations efficiently. Compared with the traditional methods, the momentum-based equation can estimate all the inertial parameters of the base and the target simultaneously at each steps, while the force-based equation does not require torque of the joints. To verify the validity and feasibility of the proposed methods, 2D and 3D models with different targets and initial velocities are simulated and analyzed. The results show that all the estimated values show convergence to their ideal values and the method can be easily achieved online via recursive techniques.