Designing effective control systems for Remotely Operated Underwater Vehicles (ROVs) presents significant challenges stemming from the marine environment and the intrinsic properties of the vehicle dynamics. These challenges include the ROV's highly coupled nonlinear dynamics, substantial parametric uncertainty (such as added mass and hydrodynamic coefficients), and considerable external disturbances like ocean currents and wave interactions. A critical practical constraint in ROV control is actuator saturation, as physical thrusters impose limits on the maximum output force and torque. Since standard control design methodologies, including the conventional SDRE technique, typically assume unbounded control input, failing to account for these physical constraints can decisively degrade performance and potentially lead to system instability. The SDRE method conceptually extends the Linear Quadratic Regulator (LQR) framework by transforming the nonlinear system dynamics into a pseudo-linear structure using State-Dependent Coefficient (SDC) matrices. This approach allows for the computation of a suboptimal control law by solving the State-Dependent Algebraic Riccati Equation (SDRE), thereby minimizing a quadratic-like performance index over an infinite time horizon. The objective of this research is to design a constrained SDRE controller for ROVs that explicitly addresses and manages the input saturation constraints required for safe and practical ROV operation.
