Nonlinear model predictive control using symbolic computation on autonomous marine surface vehicle.

Nonlinear model predictive control (NMPC) has become a preferred option particularly for controlling a strongly nonlinear system due to an increase in the availability of powerful cheap computing resources in recent years. However, there are still some hurdles to widely apply the NMPC algorithm for fast nonlinear dynamic systems. One of the primary challenges is how to quickly solve a set of nonlinear differential model equations and a nonlinear dynamic optimization problem in real time. A large portion of the computational problems arises from gradient vector and Hessian matrix derivations and evaluations. In this paper, the problem has been handled using an advanced symbolic computation tool called SymPy. With the symbolic computation, the complex gradient vector, Hessian matrix, and nonlinear dynamic optimization problem are derived analytically and evaluated with machine accuracy. As the rising need for a higher safety level, autonomous marine surface vehicles have been vastly used for various applications in recent years. It is known that model-free methods such as proportional-integral-derivative and nonlinear sliding mode controls are not sufficient to provide a robust control for a highly nonlinear surface vehicles. In this paper, the developed NMPC algorithm has been adopted to solve the two-dimensional trajectory tracking problem for an autonomous marine surface vehicle. [ABSTRACT FROM AUTHOR]