Sliding Mode Pose Control for Six-DOF Air-Bearing Testbed Using Neural Learning and Reconstructed Allocation

This paper focuses on developing a sliding mode pose control (position and attitude) strategy for a six-degree-of-freedom (6-DOF) air-bearing testbed, which is subject to model uncertainties and multivariate disturbances. To account for the complex coupling characteristics of full-degree-of-freedom motion, the 6 -DOF model is constructed based on kinematics and dynamics. To achieve effective tracking, the sliding mode pose control is designed by using the historical stack for online learning and nonlinear disturbance observer to compensate for multivariate disturbances. Additionally, a null-space closed-loop reconstructed allocation method is utilized to reconfigure the over-limit thrust of cold air thrust equipment, which ensures both optimal energy usage and minimal distribution error. The stability of the system is proved via Lyapunov analysis, and simulation results are presented to demonstrate the effectiveness of the proposed strategy.