Robust Adaptive Control of an Uninhabited Surface Vehicle

In this paper, we develop a novel and robust adaptive autopilot for uninhabited surface vehicles (USV). In practice, usually asudden change in dynamics results in aborted missions and the USV has to be rescued to avoid possible damage to other marine crafts inthe vicinity. This problem has been investigated in our innovative design, which enables the autopilot to cope well with significant changes in the system dynamics and empowers USVs to accomplish their desired missions. The model predictivecontrol technique is employed which adopts an online adaptive nature by utilising three algorithms. Even with random initialisation,significant improvements over the gradient descent and least squares approaches have been achieved by the modified weightedleast squares (WLS) method, which periodically reinitialising the covariance matrix. Extensive simulation studies have been performed to test and verify the advantages of the proposed method