Dynamic Modeling and Tube-MPC Based Landing Control for a Robotic Manta With Flexible Fins

摘要

The median and/or paired fin (MPF) propulsion mode underwater robot has gradually gained increasing attention due to its strong stability and load capacity. This article focuses on the dynamic modeling method for MPF robotic fish and its landing control. First, a more realistic three-dimensional dynamic model is established, considering the chordwise flexibility of distributed-driven flexible fins along with ideal and viscous fluid forces. Then, the model parameters are determined by less consumption and data demand method which incorporates computational fluid dynamics and a data-driven particle swarm algorithm. Besides, considering uncertain dynamics, a landing control scheme is established by combining the line-of-sight guidance law, tube-based model predictive control method, and improved central pattern generator driven network. Finally, various motions and landing simulation and experiment results confirm the established model, determined parameters, and control scheme. This article heralds a new avenue for advancing the modeling, control, and experiment associated with MPF robotic fish, particularly in landing missions under complex and varying underwater environments.