Date of Award

6-2026

Document Type

Thesis

Publisher

Santa Clara : Santa Clara University, 2026

Degree Name

Master of Science (MS)

Department

Electrical and Computer Engineering

First Advisor

Burak Kürkçü

Abstract

This thesis explores the coupling of Control Coherent Koopman (CCK) modeling and robust algebraic H→-based stable inversion to combine the mathematical rigorousness of physics-based nonlinear control with the linear control synergy of data-driven Koopman-based linear control. CCK modeling augments a physical model of a nonlinear system with physical/virtual actuator dynamics and uses Hilbert-space projections to produce a finite-dimensional linear time-invariant (LTI) Koopman model with an exact constant actuator-side input matrix and structured additive uncertainty. This uncertain LTI model is then used to design an algebraic H→-based stable inversion controller to robustly control the nonlinear system. This thesis presents theorems that explicitly define the additive uncertainty structure and prove the accuracy and robustness of algebraic control. Numerical studies are included to demonstrate the prediction accuracy of CCK modeling and the tracking accuracy and robustness of CCK-based algebraic control.

Available for download on Sunday, February 28, 2027

Share

COinS