Non-linear Path Following Control of Underactuated Autonomous Surface Vehicles

The globally K-exponentially stable path following control of an underactuated autonomous surface vehicle (ASV) is proposed based on the cascaded approach. The dynamics model of the ASV path following errors is derived utilizing the Serret-Frenet coordinate originated at the free path reference point, which creates an added control input, i.e., the change rate of the path parameter, to avoid the singularity problem when adopting the coordinate originating at the orthogonal projection point. The path following reference yaw is designed, and the whole path following model is decomposed into two cascaded sub-models, named the position tracking sub-model and the yaw angle and surge velocity tracking sub-model. The global K-exponential stability of the path following errors is proved using the cascaded system theory with the globally exponentially stable yaw angle and surge velocity tracking controllers. The results of the mathematic simulations and the lake trials of a real ASV demonstrate the performance of the proposed path following control algorithm.