Robust and Adaptive Control

Control of underwater vehicles at low speed is challenging, as the authority of the control surfaces increases with the velocity squared, and thus at low speed the control surfaces are less effective. Procedures involving propeller power bursts, that work in surface ships where rudders are installed downstream of the propellers, don’t work in underwater vehicles where tail planes are located upstream. Using larger control surfaces can alleviate the problem, at the cost of higher resistance in forward operation. Operation near the surface results in additional layers of complexity to the problem. The presence of the free surface changes the hydrodynamics of the vehicle, reducing the effectiveness of the sail and top control surfaces and producing a suction force that attracts the submarine to the surface and produces a bow down pitching moment. These effects complicate depth-keeping and depth-change controllers. 

Together with researchers at the IIHR-Hydroscience & Engineering at University of Iowa, we focus on the development, implementation and validation of nonlinear robust and adaptive controllers for underwater vehicles that exhibit guaranteed performance in challenging scenarios. Computational Fluid Dynamics (CFD) simulations is employed to derive vehicle dynamic models. The primary geometry that we consider is the generic submarine Joubert BB2 (Figure 1), which is used as a basis to design and evaluate the control algorithms. Nonlinear controllers are designed to exhibit optimality and robustness properties in the presence of unmodeled dynamics, uncertainties, and disturbance. Particular emphasis is placed on controller design for depth change and controlled turn maneuvers in two particularly adverse conditions, namely near-surface operation and low-speed maneuvering. 

The controllers are designed and developed by combining efforts in approximation theory and direct methods for optimal control, as well as robust and adaptive control theory. The solutions will be tested using the code REX (see Figures 2 and 3), developed at The University of Iowa, which offers a unique platform to perform advanced evaluation of the controller’s performance, by resolving the physics of the hydrodynamics and body motions of the underwater vehicles under investigation while accurately implementing the proposed control architecture. 

Figure 1: Joubert BB2 geometry.
Figure 2: REX submarine maneuver computations: Joubert BB2 in a turn maneuver
near the surface
Figure 3. DARPA Suboff in a horizontal overshoot maneuver.
%d bloggers like this:
search previous next tag category expand menu location phone mail time cart zoom edit close