Neural Network-Based Adaptive Finite-Time Consensus Tracking Control for Multiple Autonomous Underwater Vehicles

Abstract

Considering the problem of consensus tracking control for multiple autonomous underwater vehicle (AUV) system, a neural network-based finite-time nonsingular fast terminal sliding mode control method is proposed. First, in order to elaborate on the communication relationship, the algebraic graph theory is combined with a leader-follower architecture. Next, the modified nonsingular fast terminal sliding mode is adopted to improve the fast response characteristic of the system, and distributed control laws are constructed based on the force analysis of each AUV. Furthermore, neural networks technique is employed to approximate the uncertain dynamics and forces caused by the harsh environment of the ocean. Finally, it is proved that the tracking errors can converge to a small neighborhood of the origin by using the proposed algorithm. The effectiveness and robustness of the proposed algorithm are illustrated by a simulation example. INDEX TERMS Adaptive neural control, multiple AUV system, nonsingular terminal sliding mode, finite time, graph theory.

Autonomous Underwater Vehicles: Localization, Navigation, and Communication for Collaborative Missions

“…Cui et al [131] focused on the problem of tracking control for multi-AUV systems and proposed an adaptive fuzzy-finite time control method. In this algorithm, algebraic graph theory is combined with a leader-follower architecture for describing the communication of the system.…”

Robust Adaptive Tracking Control of Underwater Vehicles: Design, Stability Analysis, and Experiments

“…where all the terms in (15) have been defined in (2) before transforming to R n . Theorem 1 : The proposed controller defined by (13) and the adaptation dynamics (14) ensure that the trajectory (9) of the uncertain coupled nonlinear MIMO dynamics subject to external disturbance and defined in (15) is bounded and converges exponentially to the desired trajectory (8) in a finitetime t f (and remains in (8) ∀t ≥ t f ); as long as the proposed RAC gains are selected sufficiently large enough based on the initial condition of the vehicle's states, while the parameter ρ satisfies the following conditions:…”

“…As a consequence, the design of a tracking control scheme for the such vehicles becomes a critical issue [9]. Indeed, several control schemes have been proposed in the literature to address this problem, such as improved control schemes based on PD and PID control [10] [11] [12], robust control [13], adaptive control [14], intelligent and hybrid control [15].…”

An improved swarm intelligence algorithms-based nonlinear fractional order-PID controller for a trajectory tracking of underwater vehicles

“…In [1], a model-predictive controller MPC is proposed to control depth signals using quadratic programming, and in [2] the researcher discusses a model-free reinforcement learning algorithm for the AUV. Additionally, in [3] applied a neural network for a consensus multiple tracking AUV problems, in [4] a routing protocol is proposed to solve the end-to-end delay in AUV System, and finally on [5] discussed the adaptive AUV system design. In this research work,  ISSN: 1693-6930 an nonlinear fractional order proportional integral derivative (NL-FOPID) AUV is intended to solve the unknown disturbances' problems that affect system response and compared to the relative PID controller.…”