An efficient hybrid approach for trajectory tracking control of autonomous underwater vehicles. (February 2020)

• Record Type: Journal Article
• Authors: Kumar, Naveen; Rani, Manju
• Is Part Of: Applied ocean research. Volume 95(2020)
• Publication Date: 2020-02
• Subjects: Autonomous underwater vehicle -- Model-dependent controller -- Position tracking -- Radial basis function neural network -- Lyapunov stability criteria; Ocean engineering -- Periodicals; 620.416205
• DOI: 10.1016/j.apor.2020.102053
• Languages: English
• ISSNs: 0141-1187

Abstract

Highlights: An efficient hybrid control scheme is proposed for trajectory tracking control of AUV. The control scheme integrated the benefits of model based and model free controllers. RBF neural network is used to deal with the uncertainties of the system. The effects of disturbances and reconstruction error are compensated with an adaptive compensator. The system is shown to be stable utilizing Lyapunov theory. Simulation studies are performed to show the effectiveness in a comparative manner. Abstract: In this manuscript, an efficient hybrid trajectory tracking control scheme has been proposed for an autonomous underwater vehicle in the presence of structured and unstructured uncertainties. To cope with these uncertainties, the model-dependent control scheme is successfully combined with the model-free control scheme. Due to the effects of the uncertainties, the full knowledge of the dynamic model of the vehicle cannot be accurately obtained in real applications. Therefore whatever the partial information is available about the dynamics of the system has been utilized in the controller design. A radial basis function neural network is utilized for the approximation of the unknown dynamics without requiring offline learning. To compensate for the unknown effects like the external disturbances and the reconstruction error of the neural network, an adaptive compensator is also added to the part of the controller. For the stability analysis, the online learning of the parameters and the neural network weights are used in the Lyapunov approach. Based on the Lyapunov stability criteria and Barbalat lemma, the tracking errors converge to zero asymptotically. Finally, comparative numerical simulations are performed over a four-degree of freedom autonomous underwater vehicle and efficiency and applicability of the proposed control framework is validated in a comparative manner with the existing controllers.