Probability-oriented disturbance estimation-triggered control via collaborative and adaptive Bayesian optimization for reentry vehicles

The paper investigates the performance improvement issue for reentry vehicles under uncertainties from the perspective of probability. The disturbance estimation-triggered control (DETC) proves to achieve transient performance increase compared with the standard disturbance-observer control methods, and the presented approach further exploits the probability-oriented transient performance improvement based on the collaborative and adaptive Bayesian optimization (CABO) technique, which constructs the main contribution of the paper. Based on the attitude dynamics of reentry vehicles, the DETC method is first introduced to guarantee the tracking stability and robustness against the uncertainties including the aerodynamic perturbation and wind effects. Meanwhile, the performance improvement is analyzed theoretically. Then, by virtue of the CABO algorithm, the CABO-based DETC is presented by combining the performance and probability indexes. Finally, the simulation results verify the effectiveness of the proposed control scheme and parameters influence is also discussed.