Oceanic Dynamic Positioning through Hierarchical LQR Algorithm-Driven Advancements for Underactuated Remotely Operated Underwater Vehicles

Marine biodiversity mapping is crucial for conservation efforts. However, the dynamic underwater environment poses challenges in achieving accurate and stable mapping using remotely operated vehicles (ROVs). This research evaluates three control algorithms, proportional-integral-derivative (PID), model predictive (MPC), and linear quadratic regulator (LQR), for ROV dynamic positioning through simulations. Using a mathematical model of the ROV, the study tests position-hold capabilities and disturbance rejection. The LQR outperforms the other methods, demonstrating rapid response times and effective disturbance recovery. In contrast, MPC shows degraded tracking, and PID exhibits sluggish responses. Introducing a multi-stage control architecture with an inner LQR loop further enhances trajectory tracking. The findings suggest that the LQR, especially when optimized and adaptive, provides superior dynamic positioning for ROVs in maritime applications. Future work aims to implement this control approach in a real-world ROV system.