Guidance, Navigation and Control

To guide multi-agent systems (MASs) through a cluttered environment, this paper proposes a distributed control method based on a novel fractional-order extended state observer (NFO-ESO). First, unlike the traditional multi-agent double-integral model, the agent model constructed in this paper contains external wind disturbances (mismatched disturbances) and internal unmodeled dynamics (matched disturbances), and these disturbances are treated as matched disturbance equivalent. Then, the designed disturbance observer is added to the robust distributed controller to coordinate the disturbance, which is used to accomplish the curve virtual tube crossing task in the perturbed case. The disturbance observer is NFO-ESO, which is implemented in this paper by constructing a neural network based on fractional-order discrete theory. It improves the weakness of the traditional ESO that only observes low-frequency slow time-varying disturbances and can be more suitable for nonlinear fast time-varying disturbance estimation. Finally, to demonstrate the advantages of the proposed controller, we show an example of MASs with wind disturbance and unmodeled dynamic through numerical simulation experiment. The instance shows the effectiveness of the proposed distributed control method under external disturbances.

This paper addresses the problem of synthesizing an anti-windup compensator for discrete variable delay systems with missing measurements. The descriptor redundancy approach, embedding the Lyapunov–Krasovskii functional, relaxed summation inequality, and linear matrix inequality (LMI) techniques are applied to guarantee the stability of the closed-loop system. An optimization algorithm is proposed to maximize the estimated domain of attraction that can be easily implemented. Numerical examples are given to illustrate the advantages of the proposed procedure over other methods available in the literature and to show that it is less conservative than other approaches.

Modern beyond visual range air game is characterized by high dynamic complexity and multi-factor interweaving, making it challenging for traditional decision-making methods. This paper constructs an autonomous decision-making framework for a single-plane beyond visual range air game that encompasses the entire game process. We design tactical rules with both offensive and defensive capabilities and formulate an autonomous air game decision-making method. This method is primarily driven by rules and is complemented by an intelligent maneuver decision-making approach based on the Double Deep Q-Network (DDQN). Simulation results demonstrate that our designed method effectively enhances the maneuvering decision-making ability of aircraft and has promising application prospects.

Investigating latent interactions beyond direct connections is essential for analyzing complex networks. However, traditional graph structures often fail to capture complex relationships, especially in the high-order interactions among multiple individuals. To address this issue, we extend the graph isomorphism network (GIN) framework to hypergraphs, treat nodes as self-hyperedges, and propose the self-hypergraph isomorphism network (SHGIN). Meanwhile, the hypergraph Weisfeiler–Lehman (WL) test is also proposed to distinguish different isomorphisms of hypergraphs and improve the representation power of hypergraph neural networks. Extensive experiments on co-authorship and co-citation networks demonstrate the effectiveness of SHGIN. The results indicate that our model displays superior hypernode classification accuracy compared to traditional graph neural networks in semi-supervised learning (SSL). Furthermore, it surpasses existing hypergraph neural network models in co-authorship datasets, highlighting its effectiveness in capturing high-order relationships in complex networks.

The control performance of model-based control approaches depends on the accuracy of the system dynamics. However, due to unknown environmental disturbances, quadrotor dynamics are time-varying during real-world flight, and real-time updating of the dynamics is required to avoid degradation of control performance due to model errors. To address this challenging issue, this paper proposes a hybrid controller that combines model predictive control (MPC) with model-based reinforcement learning (MBRL) to improve quadrotor trajectory tracking accuracy in unknown wind conditions. The MPC controller uses simple dynamics identified from offline flight data. The MBRL controller learns MPC-augmented dynamics represented by the Gaussian process (GP) and the residual policy from online flight data. The RL controller outputs additional actions to compensate for MPC control errors, while MPC can ensure the safety of RL exploration. The computational cost of GP when dealing with a large dataset affects the real-time performance. We design a priority data selection criterion to maintain a small dataset, balancing model accuracy and computational time. The accurate tracking results for different reference trajectories under various wind conditions demonstrate the robustness of the proposed approach. Comparative results with nominal MPC and related baselines show the advantages of the proposed approach in terms of control and real-time performance.

This paper mainly studies the electrical current control of the spinning bare electrodynamic tether system (SBEDT) during its spin-up process. The SBEDT combines the high electrical absorption efficiency of bare tethers and the centrifugal stability of spinning tethers, which provides a propellant-free solution to missions including space debris removal, artificial gravity, and payload transportation. However, the nonlinear current distribution of bare tethers poses serious challenges to its dynamic and control. To deal with this problem, this paper focuses on the current analysis and optimal control of tether motions. First, a dynamic model is established for analyzing the varying currents in different positions, and the analytical expressions of the electrodynamic force are given. Second, the minimum tether length for spin-up is calculated by the maximum equilibrium position, providing a theoretical basis for system design. Third, an optimal tether deploying/retrieving controller is proposed to adjust electrodynamic force. The control law is designed according to the Bellman optimality principle, which tracks the angular velocity and controls the tether length within a certain range. Under the proposed analyzing method and the corresponding control strategy, spinning stability can be maintained during the spin-up process and the system successfully reaches the desired spinning state.

Navigating in an environment, where Global Navigation Satellite System (GNSS) signals are unavailable, presents a formidable challenge. A feasible strategy involves integrating stellar observations with inertial information to realize precise navigation. Inertial/Stellar Celestial Navigation enhances position accuracy by addressing gyro drift through stellar observations. However, the position error still continues to grow over time which arises from unobservable error factors. With improved orbit determination accuracy and the resident space object (RSO) observation method, navigating using RSO observation is gaining viability. This study models the RSO navigation problem as a Perspective-n-Point (PnP) problem and proposes that observing more than 6 RSOs can yield complete solution for position and attitude. Additionally, a tightly coupled navigation algorithm integrating inertial, stellar, and RSO observations is introduced. Simulation validation under dynamic condition demonstrates the algorithm’s efficacy, achieving a position accuracy of 30m and an attitude accuracy of 5arcsec (RMS) without GNSS data.

Control moment gyroscopes (CMGs) are a favorable choice for spacecraft attitude control systems thanks to their torque amplification capability. However, their performance is hindered by geometric singularities, which can severely limit control capabilities during attitude maneuvers. This paper proposes a hybrid attitude control system incorporating a single reaction wheel into the standard minimum redundant four-CMG pyramid configuration, providing an additional degree-of-freedom to avoid and escape singularities effectively. A comprehensive mathematical analysis of the system’s Jacobian matrix, using row echelon form, is performed alongside a geometrical analysis of the hybrid configuration’s momentum envelope. These analyses demonstrate the reflection of a reaction wheel inclusion into the system of four CMGs in the improvement of the system’s capacity to avoid/escape singularity. Additionally, an asymptotic control law and an adapted steering law are developed to optimize control performance. The effectiveness of the proposed hybrid system is validated through simulation, which includes a comparative analysis with traditional CMG-only system. The results highlight the superiority of the hybrid system in handling singularities.

In this paper, for the two/three targets in the circle orbit with the same altitude, a numerical method is presented to design an orbit for a spacecraft to visit multiple targets without maneuvering. First, for the two targets’ case, an equation is established by the relationship of the two angles that both are the function of the time, one of the angles is the angle between two targets’ geocentric position vectors at two times, and another one is the spacecraft flight angle during the two times. The visit time can be solved by the assumption that one of the visit times is known in the equation, and the orbit element can be transferred from the position and velocity vector determined by the solved visit time. Then, for the three targets’ case, the problem is decoupled into two cases of two targets, and there is a public target in each case. Based on the method for two targets, two different orbits can be obtained by determining the visit time of the public target. The difference between two orbits can be expressed by the difference in velocity vectors of the public target at different times. A sampling method is utilized to select two velocity vectors with relatively small differences as initial values, the Newton iteration method is used to obtain visit times with higher accuracy. Simulation results illustrate the effectiveness of the orbit design methods.

This study focuses on enhancing the evasion capabilities of unmanned ground vehicles (UGVs) using Generative Adversarial Imitation Learning (GAIL). The UGVs are trained to evade unmanned aerial vehicles (UAVs). A decision-making neural network has been trained via GAIL to refine evasion strategies with expert demonstrations. The simulation environment was developed with OpenAI Gym and calibrated with real-world data for the improvement of accuracy. The integrated platform including the proposed algorithm was tested in flight experiments. Results showed that the UGVs could effectively evade UAVs in the complex and dynamic environment.