QUASILINEARIZATION, SYSTEM IDENTIFICATION, AND PREDICTION
Abstract
An adaptive controller is one which has the capability of learning about unknown aspects of a system being controlled and then modifying its control regime in an effort to improve the quality of the control exerted. A mathematical formulation and computational solution of the problems of system identification and the determination of unmeasurable state variables on the basis of observations of a process, two topics of central importance in the design of adaptive controllers are presented. The approach suggested-based on the theory of quasilinearization-is an outgrowth of continuing RAND research on the computational solution of multi-point boundary-value problems. The paper should be of interest to control engineers and numerical analysts.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 01, 1963
- Accession Number
- AD0414525
Entities
People
- Harriet H. Kagiwada
- Richard E. Bellman
- Robert E. Kalaba
Organizations
- RAND Corporation