An Adaptive Parameter Identification Algorithm for Post-capture of a Tumbling Target

Abstract

In this paper, a new parameter identification scheme is proposed for an unknown tumbling target captured by the space manipulator. Due to the unknown and various dynamic parameters of the target and the strong nonlinear characteristics of the manipulator, it is difficult to achieve the capture operation with high tracking accuracy and low energy consumption. Aiming at the challenge, the parameter identification model of free-floating space manipulator is established based on the momentum conservation principle. Then, the VDW-RLS (variable data-window-size recursive least square) algorithm is applied to identify the inertia parameters. VDW-RLS algorithm can adjust the size of data window online according to the change of system parameters. SimMechanics simulation platform is employed to implement the three-dimensional model of 7-DOFs manipulator system. Simulation results show that the proposed algorithm has a fast tracking performance and low misalignment in the inertial parameter identification of the multi-DOFs space manipulator, and the control effect can be further improved by using the identified dynamic parameters.