Filtered finite-time ESO-based prescribed performance control design with application to test mass capture of drag-free spacecraft

In this paper, the prescribed performance control problem is mainly investigated for nonlinear systems that simultaneously consider unknown disturbances and measurement noise. First, a new filtered finite-time extended state observer (FFTESO) is presented, and the designed variable-order filter and observer properly integrate the fractional power term such that FFTESO can accurately estimate the system state and disturbance within a finite time, dealing with the trade-off between finite-time convergence and sensitivity to measurement noise. Next, the FFTESO-based output-feedback prescribed performance control is proposed by utilizing the barrier Lyapunov function and the flexible prescribed performance function to ensure that the tracking error remains within predefined boundary, while effectively addressing actuator saturation constraints. In addition, a comprehensive numerical comparison of the observer performance is conducted to demonstrate the advantages of the proposed FFTESO in noise suppression and fast convergence. Finally, the proposed control scheme is applied to the test mass capture of drag-free spacecraft, which requires strict constraints on the attitude and displacement of the test mass, and the hardware-in-the-loop experiments with Monte Carlo tests under worst-case release conditions clearly demonstrate that the proposed control scheme can achieve safe capture of the test mass to avoid collisions between the test mass and the spacecraft cavity.