Model Predictive Control Using State Estimation Based on Unscented Kalman Filter for Stabilization of Underwater Vehicle Dynamics

This research addresses the challenge of designing control systems to stabilize the nonlinear dynamics of underwater vehicles. Model Predictive Control (MPC) is a widely recognized technique that determines the current control input by solving an optimal control problem. However, MPC cannot be directly applied to systems where all state variables are not precisely known. Typically, state variables are inferred from sensor measurements, meaning that only a subset of them is available for control input design. This study aims to develop a control approach that stabilizes underwater vehicle dynamics by integrating a state estimation method into the MPC method. The novelty of this study is to develop a control framework that integrates MPC method with the state estimation method based on Unscented Kalman Filter (UKF) for stabilization of underwater vehicle dynamics.