IDENTIFICATION OF LINEAR SYSTEMS VIA NUMERICAL INVERSION OF LAPLACE TRANSFORMS,
Abstract
A fundamental problem in adaptive control, and elsewhere in science, is that of identifying the parameters of a system on the basis of observations of a process which the system undergoes. The quasilinearization technique provides a general method for doing this, whether the system is linear or not. This Memorandum introduces a new method for identifying linear systems based upon the ability to invert Laplace transforms numerically. The method employs irregularly spaced observations and is extremely fast. Results of a test calculation for a second-order linear system are given. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1964
- Accession Number
- AD0444517
Entities
People
- Harriet H. Kagiwada
- Richard E. Bellman
- Robert E. Kalaba
Organizations
- RAND Corporation