Monday, May 20
4:50-6:50 PM
Saanich 1
MS11
Optimization and Differential-Algebraic Equations (Part I of II)
(Invited Minisymposium)
During the last decade, the optimization of systems involving differential algebraic equations
has become an important research area in applications involving chemical processing, robotics,
structural analysis, and aerospace engineering. Frequently, the problem may originate as an
optimal control problem where a performance index is to be maximized, or as a prescribed path
control problem. Usually, the problems have multiple phases (with possibly unknown junction
times of the phases) across which the dynamic equations may change or the inequality path
constraints may become active or inactive. Many numerical techniques have evolved for the
solution of these problems. Typically, the numerical solution of the continuous problem is
determined by a discrete optimization technique, and it is important to understand the
connections between the continuous solutions and the discrete solutions. In this minisymposium,
the speakers will discuss various optimization based approaches to solving these problems,
including multiple shooting and direct transcription (or collocation) techniques for solving
optimal control problems, parallel methods for large scale systems, and optimization of the mesh.
Organizers: Kathryn E. Brenan, The Aerospace Corporation; and
John T. Betts, The Boeing Company
- Parallel Numerical Methods and Software for Optimization and Control of Large-Scale
Differential-Algebraic Systems
- Philip Gill, University of California, San Diego; Linda R. Petzold and Ben Rosen,
University of Minnesota, Minneapolis
- Aspects of Differential-Algebraic-Inequality Solver and Applications
- Uri M. Ascher, Ray Spiteri, and Dinesh K. Pai, University of British Columbia, Canada
- Higher Index DAE in Trajectory Optimization
- V. H. Schultz, H. G. Bock, and M. C. Steinbach, Universitat Heidelberg, Germany
- A Bilevel Framework for Dynamic Process Optimization
- P. Tanartkit and L. T. Biegler, Carnegie Mellon University
LMH, 3/15/96