A Study on the Development of a Predictive-Based Integrated Control System for Station-Keeping of Underwater Drones in Current Environments

This study proposes a predictive integrated control system that combines sensor fusion based on Invariant Extended Kalman Filter(InEKF) and Model Predictive Control(MPC). To verify the performance of the proposed system, a MATLAB/Simulink-based six-degree-of-freedom simulator was constructed and evaluated over six scenarios(current 0.3 to 0.8 m/s, depth 50 to 100 m, uniform/time change/turbulent current). The proposed framework integrates sensor fusion, predictive control, and disturbance compensation, resulting in a 77.6% improvement in positioning accuracy and a 44.1% reduction in energy consumption compared to PID control. It has high commercialization potential using commercial sensors and general-purpose hardware, and can be applied to various marine applications such as submarine cable inspection, marine observatory maintenance, and underwater archaeological investigation.