Summary
Suppose that a system undergoes a process described by a set of differential equations. The equations contain some unknown parameters and not all the initial conditions are known. Observations are made on some of the state variables during the course of the process. We wish to determine the system parameters and initial conditions which lead to best agreement with the observations.
Inverse problems of the type sketched, which are important throughout mathematical physics and engineering, are cast in mathematical form as nonlinear multipoint boundary value problems. Then computational solutions via quasilinearization and differential approximation are suggested.
Zusammenfassung
Angenommen, ein System werde einem Prozeß unterworfen, der durch ein System von Differentialgleichungen beschrieben wird. Die Gleichungen enthalten einige unbekannte Parameter und weiters sind nicht alle Anfangsbedingungen bekannt. Das Verhalten einiger Zustandsvariablen wird während des Prozesses beobachtet. Wir möchten nun die Parameter des Systems und die Anfangsbedingungen bestimmen, die die beste Übereinstimmung mit den Beobachtungen liefern.
Dei zu dem skizzierten Typ inversen Probleme, die in der mathematischen Physik und im Ingenieurwesen wichtig sind, werden als nichtlineare Randwertaufgaben mit Bedingungen an mehreren Stellen mathematisch formuliert. Schließlich werden für rechnerische Lösungen Quasilinearisierung und Differentialapproximation vorgeschlagen.
Similar content being viewed by others
References
Zadeh, L. A.: “On the Definition of Adaptivity,” Proc. IEEE51, No. 3, 469 (1963).
Gibson, J. E., K. S. Fu, J. B. Pearson, Z. V. Rekosius, andR. Sridhar: Modern Aspects of Automatic Control, Vol. 2, Chap. 7. Lafayette, Indiana/Purdue University, 1963.
Kalaba, R. E.: “Some Aspects of Quasilinearization,” a chapter in Nonlinear Differential Equations and Nonlinear Mechanics. New York/Academic Press, 1963.
Bellman, R. E., R. E. Kalaba, andB. Kotkin: Differential Approximation Applied to the Solution of Convolution Equations, RM-3601, The RAND Corporation, 1963.
Gibson, J. E.: Nonlinear Automatic Control, Chap. 11, New York/McGraw-Hill, 1963.
Bellman, R. E. andR. E. Kalaba, “Dynamic Programming and Adaptive Processes: Mathematical Foundation,” IRE Trans. on Autom. Control AC-5, 5 (1960).
Bellman, R. E., R. E. Kalaba, andH. Kagiwada, Quasilinearization System Identification and Prediction, RM 3812, The RAND Corporation, 1963.
Bellman, R. E. andR. E. Kalaba, Quasilinearization and Nonlinear Boundary Value Problems. New York/American Elsevier Publishing Co. (1965).
Bellman, R. E., R. E. Kalaba, andH. Kagiwada: “Orbit Determination as a Multi-point Boundary Value Problem and Quasilinearization,” Proc. Nat. Acad. Sci. U. S. A.48, 1327 (1962).
R. E. Kalaba: “On Nonlinear Differential Equations, the Maximum Operation and Monotone Convergence,” J. Math. Mech.8, 519 (1959).
Berkovitz, L. D.: “Variational Methods in Problems of Control and Programming,” J. Math. Anal. Appl.3, No. 1 (1961).
Strejc, V.: “Evaluation of General Signals with Non-zero Initial Conditions,” Acta Tech. Acad. Sci. Hungar.6, 378 (1961).
Author information
Authors and Affiliations
Additional information
With 2 Figures
This research is sponsored by the United States Air Force under Project RAND-contract No. AF 49 (638)-700.
Rights and permissions
About this article
Cite this article
Bellman, R., Kalaba, R. & Sridhar, R. Adaptive control via quasilinearization and differential approximation. Computing 1, 8–17 (1966). https://doi.org/10.1007/BF02235849
Received:
Issue Date:
DOI: https://doi.org/10.1007/BF02235849