Iterative Learning-Based Path Optimization With Application to Marine Hydrokinetic Energy Systems

摘要

This article presents an iterative learning control (ILC)-based approach for optimizing the flight path geometry of a tethered marine hydrokinetic (MHK) energy system. This type of system, which replaces the tower of a conventional system with a tether and a lifting body, can capture energy either through an on-board rotor or by driving a generator with tension in the tether. In the latter mode of operation, which represents the focal point of this effort, net positive energy is generated over one cycle of high-tension spool-out followed by low-tension spool-in. Because the net energy generation is sensitive to the shape of the flown path, we employ an iterative learning update law to adapt the path shape from one lap to the next. This update law is complemented with an iterative power take-off (PTO) controller, which adjusts the spooling profile at each iteration to ensure zero net spooling. We present and validate the proposed control approach in both uniform and spatiotemporally varying turbulent flow environments, based on a realistic ocean model detailed in this article. Finally, based on simulation results across a wide range of excitation levels, we perform a simulation-based assessment of convergence properties, comparing these results against bounds derived in the authors' prior work.