An Efficient Diving Control Scheme for Autonomous Underwater Vehicle Experiencing Actuator Failure: A Multi-Layer Cubature Kalman Filter-Based Nonlinear Model Predictive Control Approach

Autonomous underwater vehicles (AUVs) are primarily capable of withstanding prolonged periods in unstructured marine environments. These are subject to several uncertainties, including the effects of underwater currents or drift, topography, water pressure, and other factors. The inherent motion of the vehicle is influenced by buoyancy or mass error and other slowly varying hydrodynamic coefficients. Moreover, an AUV faces various types of undetected faults during its mission execution. The use of large-dimensional, rigorous onboard sensor data from real-time AUV operations, subject to the limited power capacity and the presence of other acoustic issues, makes it more challenging for the control community from both stability and positioning accuracy perspectives. Hence, an anti-disturbance stabilized control mechanism is intended to perform a variety of underwater tasks in various uncertain oceanic environments. A two-stage high-degree cubature information filter (TSHDCIF) is designed to effectively observe unmatched external disturbances, uncertain hydrodynamic parameters, and actuator faultiness. A nonlinear model predictive control (NMPC) scheme incorporated with the TSHDCIF observer is proposed in this work. Various realistic scenario-based result sets, including 1) effects on ocean currents, 2) variations in hydrodynamic parameters, and 3) loss of actuator effectiveness, are addressed in this study. The control solution is compared with the baseline NMPC scheme to examine its efficacy, including depth accuracy, control aggression, convergence analysis, and computation time per iteration. Based on the analysis, the TSHDCIF-NMPC scheme outperforms the traditional NMPC scheme by 5.49% in terms of RMSE (root mean square error) and 1.73% in terms of mean square deviation (MSD). The stability of the closed-loop system is explored using the Lyapunov theorem. Thus, it is recommended to use the TSHDCIF-NMPC scheme for systems such as AUV for better efficacy, as mentioned above.