22 March 2021 in IEEE Transactions on Circuits and Systems II: Express Briefs
This brief addresses the challenging problem of predefined-time attitude stabilization of receiver aircraft with uncertain inertia and unknown external disturbance. A sliding manifold is first designed to ensure that the sliding mode phase can be achieved in the predefined time. After that, a robust stabilization controller is developed to guarantee that attitude error falls into a small neighbourhood containing the origin with predefined-time convergence. In comparison with the conventional fixed-time and the finite-time control strategies, the key feature of this scheme is that system settling time can be arbitrarily and explicitly chosen in advance and independent of system states. Moreover, some modifications to the existing predefined-time stable system are achieved, meanwhile, the practical predefined-time stability with rigorous proof is derived. Finally, simulation results are presented to demonstrate the effectiveness of the proposed control strategy.
1 April 2021 in Aerospace Science and Technology
1 January 2022 in Space: Science & Technology
Probabilistic swarm guidance enables autonomous microsatellites to generate their individual trajectories independently so that the entire swarm converges to the desired distribution shape. However, it is essential to avoid crowding for reducing the possibility of collisions between microsatellites. To determine the collision-free guidance trajectory of each microsatellite from the current position to the target space, a collision avoidance algorithm is necessary. A synthesis method is proposed that generate the collision avoidance trajectories. The idea is that the trajectory planning is divided into macro-planning and micro-planning; macro-planning guides where the microsatellites move step by step from the initial cube to the target cube by probabilistic swarm guidance with Centroidal Voronoi tessellation, while the micro-planning is to generate the optimal path for each step and finally reach the specified position in the target cube by model predictive control. Simulation results are presented for the collision-free guidance trajectory of microsatellites to verify the benefits of this planning scheme.
17 February 2022 in Transactions of the Institute of Measurement and Control
This paper investigates the challenging attitude stabilization control problem of receiver aircraft subject to uncertain and time-varying inertia, and wind disturbances in the aerial refueling task. An adaptive fixed-time control strategy is proposed. In this approach, a predefined-time sliding manifold is preliminarily developed. The main feature is that the settling time is not only independent of initial conditions but also can be simply and explicitly defined in advance through a single parameter in the sliding mode phase. On this basis, an adaptive control law is developed to guarantee that the trajectory of attitude can be stabilized into a small region containing the origin. The upper bound of unknown disturbance is not required to implement the proposed controller, which relaxes the restrictions of design for control systems. Moreover, the measuring noises are explicitly considered and addressed to make this approach still available in the noisy measurement environment. The Lyapunov stability analysis and numerical simulations verify the effectiveness of the proposed control strategy.
11 September 2024 in IEEE Transactions on Instrumentation and Measurement
The growing significance of aerial refueling requires that receiver aircraft could perform autonomous aerial refueling tasks in flight. In this regard, precise docking is a key but challenging issue. To address this problem, a drogue trajectory prediction method based on transfer learning and attention mechanism is proposed in this article. The long short-term memory (LSTM) neural network is introduced as the base to learn temporal correlations between time series trajectory data of a drogue. To further boost the network performance, transfer learning strategy and the attention mechanism are involved in the model construction. Prior knowledge about physical models in similar domains can be passed to the network through transfer learning, and larger weights can be adaptively assigned to more important features. The effectiveness of the proposed method is verified through the comparisons with ARIMA, RNN, LSTM, and attention-based LSTM models, while effects of transfer learning and attention mechanism are visualized. When implementing this approach to perform predictive docking, a high success rate is achieved in the ground experiment, which shows great potentials for industrial applications.