Computational Geometric Mechanics and its Applications to Geometric Control Theory

 

Professor Melvin Leok

Department of Mathematics

University of Michigan

 

The geometric approach to mechanics serves as the theoretical underpinning of innovative control methodologies in geometric control theory. These techniques allow the attitude of satellites to be controlled using changes in its shape, as opposed to chemical propulsion, and are the basis for understanding the ability of a falling cat to always land on its feet, even when released in an inverted orientation.

 

Curiously, while the geometric structure of mechanical systems plays a critical role in the construction of geometric control algorithms, these algorithms have typically been implemented using numerical schemes that ignore the underlying geometry.

 

Geometric integration is the field of numerical analysis that focuses on developing geometric structure-preserving integrators, and computational geometric mechanics focuses on developing geometric integrators for dynamical systems arising from mechanics.

 

We will discuss the application of structure-preserving numerical schemes to the control of the 3D pendulum system, and more generally, the applications of discrete mechanics and geometry to the discretization of optimal control problems.

 

In addition, a discrete analogue of the method of controlled Lagrangians will be introduced, and the method is applied to the construction of a digital, real-time, feedback controller that stabilizes the inverted relative equilibrium of the cart-pendulum system.

 

This is joint work with Anthony Bloch (Math, UM), Mathieu Desbrun (CS, Caltech), Anil Hirani (CS, UIUC), Islam Hussein (Aero, UM), Taeyoung Lee (Aero, UM), Jerrold Marsden (CDS, Caltech), N. Harris McClamroch (Aero, UM), Amit Sanyal (MAE, ASU), Alan Weinstein (Math, Berkeley), and Dmitry Zenkov (Math, NCSU).

The research has been supported in part by NSF grant DMS-0504747, and a Rackham faculty fellowship and grant from the University of Michigan.

 

Friday, October 7, 2005

3:30 – 4:30 p.m.

1500 EECS