Lumped hydrodynamics identification-based cascade control for vertical-plane tracking of a fin-driven autonomous underwater vehicle

This paper aims to achieve a simplified and effective vertical-plane tracking control for a kind of fin-driven under-actuated autonomous underwater vehicles (AUVs). To this end, a two-layer framework of offline identification and online control is constructed and implemented for depth-pitch coupled tracking control of an under-actuated AUV at a constant forward speed. A simplified three-order identification model for pitch dynamics is derived and then a dedicated recursive weighted least squares algorithm is used to complete the offline estimation of lumped hydrodynamics . Subsequently, relying on identification results, an online cascade controller with just two gains and without any adaptive estimation is proposed to track the pitch guidance angle and it is proven to be input-to-state stable. Finally, comparative simulation results illustrate the effectiveness and outperformance of this two-layer framework for vertical-plane tracking control of fin-driven under-actuated AUVs.