Approximation, control and observation of complex dynamical systems
Author(s)
Gong, Zilong
Type
Thesis
Abstract
The present work contributes to the topics of approximation, control, and observation of complex dynamical systems. The thesis consists of three parts. The first part of the thesis considers the approximation of large-scale systems. First, the data-driven model order reduction (MOR) by moment-matching is considered, with a focus on multiple-input multiple-output (MIMO) systems. Then, the developed method is validated on a combined system of a 200-turbine wind farm interconnected to the IEEE 14-bus system. The second part of the thesis deals with three classes of state-feedback control problems for stochastic nonlinear systems. Exploiting a dynamic extension, approximate solutions are provided for optimal control problems and min-max zero-sum games while exact solutions are given for the H-infinity control problem. The level of approximation can be exactly quantified in terms of an additional cost, which gives the distance from the optimal solution. Finally, the observation of stochastic linear systems is presented in the last part of the thesis. An a posteriori method is considered to approximate the variations of the Brownian motion. Based on this approximation, a hybrid observer is constructed, with the property that the estimation error converges to zero asymptotically in probability as the sampling period tends to zero.
Version
Open Access
Date Issued
2024-10-22
Date Awarded
01/02/2025
License URL
Advisor
Scarciotti, Giordano
Publisher Department
Electrical and Electronic Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)