Model Order Reduction Using the Balanced State Representation: Theory, Application, and Interactive Software Implementation.
Abstract
This report investigates a model order reduction technique developed by Dr Bruce Moore of the University of Toronto, as applied to the B-52E flutter control problem currently under study by the Air Force Flight Dynamics Laboratory. The algorithm, which is based upon singular value analysis, is applied to the full twenty-fourth order model yielding an internally balanced representation which is balanced with respect to controllability and observability properties. The system is reduced in order, and comparisons are made between the Moore algorithm model and that obtained via a method used by the Air Force Flight Dynamics Laboratory. In addition to this investigation, an interactive program is presented which contains the model order reduction algorithm. Other capabilities include: estimation of the condition number with respect to inversion, singular value and condition number plotting vs. sample time for discrete time controllability, observability, and Hankel matrices, frequency response generation, and various special coordinate system transformation. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1979
- Accession Number
- ADA080371
Entities
People
- James Robert Mcclendon
Organizations
- Air Force Institute of Technology