Browsing by Author "Xin, Ming"

Now showing 1 - 2 of 2
  • Thumbnail Image
    ItemOpen Access
    Incremental twisting fault tolerant control for hypersonic vehicles with partial model knowledge
    (IEEE, 2021-05-14) Han, Tuo; Hu, Qinglei; Shin, Hyosang; Tsourdos, Antonios; Xin, Ming
    A passive fault tolerant control scheme is proposed for the full reentry trajectory tracking of a hypersonic vehicle in the presence of modelling uncertainties, external disturbances, and actuator faults. To achieve this goal, the attitude error dynamics with relative degree two is formulated first by ignoring the nonlinearities induced by the translational motions. Then, a multivariable twisting controller is developed as a benchmark to ensure the precise tracking task. Theoretical analysis with the Lyapunov method proves that the attitude tracking error and its first-order derivative can simultaneously converge to the origin exponentially. To depend less on the model knowledge and reduce the system uncertainties, an incremental twisting fault tolerant controller is derived based on the incremental nonlinear dynamic inversion control and the predesigned twisting controller. Notably, the proposed controller is user friendly in that only fixed gains and partial model knowledge are required.
  • Thumbnail Image
    ItemOpen Access
    Sensor-based robust incremental three-dimensional guidance law with terminal angle constraint
    (AIAA, 2021-07-28) Han, Tuo; Hu, Qinglei; Shin, Hyosang; Tsourdos, Antonios; Xin, Ming
    In this work, a robust incremental three-dimensional (3D) guidance law is proposed considering terminal angle constraint against maneuvering targets. As a stepping stone, the line-of-sight (LOS) tracking error dynamics is employed for the 3D guidance law design. A sliding variable is constructed such that its first-order derivative excludes the relative range in the perturbation, which avoids the unboundedness of system perturbation induced by target maneuvers near collision. A time-varying version of the sliding variable is designed to accelerate convergence of the LOS tracking errors and avoid large initial sliding variables. Then, two guidance laws are derived as a benchmark via the nonlinear dynamic inversion (NDI)-based sliding mode control (NDI-SMC) and NDI-based time-varying sliding mode control (NDI-TVSMC), respectively. To further improve guidance robustness with reduced system perturbation, the sensor-based incremental nonlinear dynamic inversion (INDI) control is used to design the INDI-SMC-based and INDI-TVSMC-based guidance laws. The sensor-based guidance laws exploit the LOS angular acceleration and guidance command output at the latest step, which result in smaller guidance gains to reject the perturbation than the NDI guidance laws. Numerical simulations in various cases and comparison studies are conducted to verify the effectiveness and robustness of the proposed method.