OPTIMAL CONTROL OF THE AC75 SAILBOAT FOR THE AMERICA'S CUP RACE

Abstract

This research focuses on the development of optimal sailing maneuvers for an AC75 foiling sailboat competing in the America's Cup. The America's Cup is the oldest, most prestigious, and technologically advanced sailboat racing competition in the world. Each iteration brings new and innovative sailboat designs which drastically improve sailing performance but increase complexity in the control of the sailboat system. This added complexity in the design and operation of the AC75 sailboat presents many challenges to the development of optimal sailing maneuvers. These challenges arise from the introduction of extra degrees of freedom and articulations in the boat such as the canting mechanisms (hydrofoils), which result in complex dynamical behaviors. The sailboat system is nonlinear, high-dimensional, and highly unstable. These complex characteristics require the development of high-order models, which are often intractable, or which introduce significant delays making them not well-suited for real-time control. The optimal maneuvers were achieved via the exploration of out-of-the-box solutions through data-driven controls and optimization. We used a high-fidelity sailboat simulator for the data generation process, and data-driven optimization schemes, such as Artificial Neural Networks (ANN), Extremum Seeking Control (ESC), and Jacobian Learning (JL) to optimize the sailing maneuvers. The optimizations were performed separately on various sailing maneuvers including close-hauled, tacking, and takeoff, as well as combinations of these maneuvers as performed during a race. The close-hauled and tacking maneuvers were optimized to achieve maximum Velocity Made Good (VMG) and minimum loss of VMG, respectively. The takeoff maneuver was optimized for maximum VMG and minimum time for the boat's transitions from displacement mode to foiling mode. The optimal solutions are subject to physical constraints and operational constraints enforced by the humans (sailors) in the loop. These maneuvers were developed for various heading angles (True Wind Angle (TWA)) and environmental conditions, such as True Wind Speed (TWS). Additionally, we performed an in-depth analysis of the optimal parameter settings obtained for close-hauled sailing to discern general trends in the parameter space. The trend of optimal parameters versus the wind direction provides a good understanding of the parameter space for varying sailing directions which can help guide the sailor's decisions during a race. The results show how optimization and controls can play a significant role in the development of advisory systems for complex human-operated systems. Lastly, the maneuvers developed in this search serve as performance benchmarks and provide insightful information about the underlying dynamics of the boat.