Back to search

National Technical University of Athens (NTUA)

Κατασκευή προσαρμοστικών αναδράσεων για μη γραμμικά συστήματα πεπερασμένης διάστασης και γραμμικά συστήματα άπειρης διάστασης

Abstract

dc:description

We present three adaptive schemes for the practical regulation of systems of ordinary differential equations with unknown parameters, appearing linearly. The schemes are designed to guarantee robustness against external disturbances. Namely, in Chapter 3 an adaptive scheme for nonlinear systems is presented which employs nonlinear damping terms that depend both on the state and the parameter estimate. We show that the scheme guarantees the existence of KL estimates along the regulation of the system state up to a point of assignable distance from the origin. Hence, the trajectories of the closed-loop system are converging uniformly, with respect to the initial conditions, to a neighborhood of the origin. The radius of that neighborhood and the convergence rate of the closed-loop system trajectories are controlled by parameters within the scheme. Thus, ensuring a degree of robustness against external disturbances. Regulation of the state to the origin is not excluded, but it happens in a non-uniform and non-quantifiable manner. Moreover, the design which is presented defines a class of output functions for which Uniform Global Asymptotic Output Stability (UGAOS) is guaranteed. We present the design applying to systems that satisfy the Matching Condition (MC) and to systems in the strict feedback form. Simulations are provided that compare the performance of the designed scheme with the classical one proposed in the literature. In Chapter 4, we present the design of an infinite-dimensional feedback that depends on the history of the state and the input function itself. This scheme, which cannot be characterized as adaptive in the classical sense, guarantees, in the presence of arbitrary large, bounded disturbances, practical regulation of the state in a uniform, with respect to the initial conditions, manner. The radius of the residual neighborhood, the convergence rate and the asymptotic gain of the disturbances are assignable. Moreover, we present an identifier that operates independently. Provided with sufficient state excitation, the identifier achieves exact estimation of the system parameters in the disturbance-free case. In the case where disturbances are present, the identifier produces an estimation with an assignable gain with respect to the magnitude of the disturbance. In Chapter 5, we extend the design of the controller Deadzone-Adapted Disturbance Suppression (DADS) for systems that satisfy the MC, to systems in the strict feedback form. The adaptive scheme Deadzone-Adapted Disturbance Suppression (DADS), guarantees, in the presence of arbitrary large, bounded disturbances, practical regulation of the state to a residual set of assignable radius. The scheme that applies to strict feedback systems is designed following the backstepping procedure and we present an explicit design algorithm for a class of strict feedback systems. Finally, a simulation is presented comparing the performance of the DADS scheme with the classical one proposed in the literature, both with the σ-modification in the adaptive law and without.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2026

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Aslanidis, Alexandros
  • Ασλανίδης, Αλέξανδρος

Subjects

dc:subject × 12

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/61260

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Aslanidis, Alexandros; Ασλανίδης, Αλέξανδρος. Κατασκευή προσαρμοστικών αναδράσεων για μη γραμμικά συστήματα πεπερασμένης διάστασης και γραμμικά συστήματα άπειρης διάστασης. National Technical University of Athens (NTUA), 2026. http://hdl.handle.net/10442/hedi/61260