Search

This article explores trajectory-following control for an unmanned aerial–aquatic vehicle (UAAV) navigating complex ocean disturbances and the interplay of air–seawater coupling factors. First, leveraging the backstepping technology, an adaptive algorithm is proposed to tackle the attitude and position following. Additionally, a nonlinear observer is crafted to estimate complex ocean disturbances. The UAAV model, characterized by six degrees of freedom (DOF) and nonlinear properties, experiences significant pose changes when emerging from water, underscoring the critical importance of precise pose control. Finally, stability analysis and numerical simulations are demonstrated to verify the feasibility and validity of the proposed control strategies. [ABSTRACT FROM AUTHOR]

This article presents an adaptive method for ArduPilot-based autopilots of fixed-wing unmanned aerial vehicles (UAVs). ArduPilot is a popular open-source unmanned vehicle software suite. We explore how to augment the PID loops embedded inside ArduPilot with a model-free adaptive control method. The adaptive augmentation, adopted for both attitude and total energy control, uses input/output data without requiring an explicit model of the UAV. The augmented architecture is tested in a software-in-the-loop UAV platform in the presence of several uncertainties (unmodeled low-level dynamics, different payloads, time-varying wind, and changing mass). The performance is measured in terms of tracking errors and control efforts of the attitude and total energy control loops. Extensive experiments with the original ArduPilot, the proposed augmentation, and alternative autopilot strategies show that the augmentation can significantly improve the performance for all payloads and wind conditions: the UAV is less affected by wind and exhibits more than 70% improved tracking, with more than 7% reduced control effort. [ABSTRACT FROM AUTHOR]

• Construction of a BiLSTM-based internal model control architecture for USV trajectory tracking: The paper establishes an architecture that incorporates a BiLSTM neural network as the internal model for trajectory tracking. It demonstrates that a model trained using BiLSTM possesses the necessary conditions for effective control, thereby addressing the nonlinear control problem in USV systems. • Addressing the challenge of online nonlinear optimization in control steps: The paper acknowledges that neural networks require solving online nonlinear optimization problems at each control step, which can be challenging in practical control applications. To overcome this, the proposed method introduces adaptive internal model control based on BiLSTM. By utilizing the computationally efficient internal model control approach, the paper effectively resolves this issue, enabling the deployment of the neural network in engineering applications. • Implementation using Raspberry Pi 3 and TensorFlow Lite, The trained model is applied in control by leveraging Raspberry Pi 3, and TensorFlow Lite. This implementation showcases the practical applicability of the proposed approach in real-world control scenarios. This paper investigates a practical adaptive internal model control (IMC) for an unmanned surface vehicle (USV) with unknown time-varying nonlinear model parameters and environmental disturbances. Firstly, an internal model of the USV is presented using the Bidirectional Long Short-Term Memory (BiLSTM) neural network. Then, an adaptive neural IMC controller is designed for the trajectory tracking of USV by using the IMC method. The internal model and the controller are updated by an error threshold algorithm. The proposed control scheme comprises of a trajectory guidance module via the Line-of-Sight (LOS) guidance method and a tracking control module designed by IMC theory. Under the proposed control scheme, the development processes of the vehicle platform and the control algorithms are described, and accurate tracking control can be achieved. Finally, the results of simulation and field experiments are presented and discussed to validate the effectiveness of the proposed control scheme. [ABSTRACT FROM AUTHOR]

In this paper, a concise leader-follower formation control approach is presented for a group of underactuated unmanned surface vehicle with dynamic system uncertainties and external environment disturbances, where the output errors are required to be within constraints. To settle the output error constraints, a standard barrier Lyapunov function (BLF) is incorporated into the backstepping control method. Furthermore, the "differential explosion" problem of virtual control laws is avoided by introducing the dynamic surface control. To estimate the unknown dynamic terms, an adaptive neural network is designed and a nonlinear disturbance observer is adopted to compensate for the approximation errors of neural network and ocean environment disturbances. Under the constraint of output error, the presented controller based on standard BLF has simpler structure and better control performance than depended on tan-type BLF. The presented controller can ensure that the formation errors converge to a small range around zero, while the output error constraint requirements are met. All signals in the closed-loop system are bounded, and the numerical simulation further shows the effectiveness of the presented control scheme. [ABSTRACT FROM AUTHOR]

This article presents a new adaptive method for formation control of unmanned aerial vehicles (UAVs) with limited leader information and communication. We study a formation control protocol in the framework of vector-field guidance where the leader can communicate its position and orientation but not its velocity. A practical motivation for this scenario is the so-called congestion-aware control, in which tradeoffs between the density of unmanned vehicles and communication interference caused by many communicating vehicles arise: these tradeoffs may require to reduce the communication load to avoid interference. To compensate for the lack of knowledge of the leader velocity, each UAV makes use of a local estimation mechanism. The resulting method is an adaptive control method, whose stability can be established using Lyapunov stability. We show that the method can be extended to a distributed communication setting with a few neighboring UAVs in place of the leader. Extensive simulations with different formation shapes (Y, V, and T formation) show that the proposed adaptation mechanism effectively achieves the formation despite the unknown leader velocity. The proposed mechanism has a very similar performance to the ideal case when the leader velocity is perfectly known, and outperforms all the nonadaptive cases in which the followers have an incorrect knowledge of the leader velocity. [ABSTRACT FROM AUTHOR]

To achieve the accurate control of the variable-mass unmanned surface vehicle (USV), this paper proposes an adaptive surge control method with uncertain variation in mass and draft. Aiming at the time-varying characteristics of the USV, the analytical expressions of mass and each hydrodynamic derivative are presented based on the draft and its higher-order terms as independent variables. The model uncertainty problem caused by quantities of nonlinear coupling terms related to the mass and draft is solved. For the high correlation between the motion state and the draft of the variable-mass USV, a sliding mode observer is designed to obtain the estimated varying mass and drat in real-time. The adaptive surge control algorithm of variable-mass USV is proposed, which can update its adaptive laws in real time based on the sliding mode observations. The overall stability of the control system is analyzed via the Lyapunov algorithm. Simulation experiments of several payload delivery missions are carried out. Results show that the proposed controller can achieve high precision surge control under typical missions for variable-mass USV, such as step variation and continuous payload change, and the control effect is remarkably better than the traditional control method. • The coupling relationships in the variable-mass USV's model are decoupled and linearized via different orders of the draft. • A sliding mode observer is designed to estimate the unknown dynamic mass and draft via the USV's motion states in 3 DOF. • The adaptive surge controller of variable-mass USV is proposed and the stability is verified through the Lyapunov algorithm. [ABSTRACT FROM AUTHOR]