Minimax Adaptive Control and Estimation
(2024)- Abstract
- This thesis presents five papers on minimax adaptive control and estimation. Minimax adaptive estimation is a framework for output prediction and state estimation that provides a priori computable performance bounds for esti- mators. Minimax adaptive controllers ensure that the closed loop has finite gain, maintaining stability and performance under model class uncertainty.
The contributions of these papers are as follows: Paper I: Presents a min- imax optimal output prediction algorithm for linear systems with parameter uncertainty. Paper II: Proposes an algorithm to compute performance bounds for minimax adaptive estimators. Paper III: Develops a minimax suboptimal adaptive controller for scalar linear systems with noisy... (More) - This thesis presents five papers on minimax adaptive control and estimation. Minimax adaptive estimation is a framework for output prediction and state estimation that provides a priori computable performance bounds for esti- mators. Minimax adaptive controllers ensure that the closed loop has finite gain, maintaining stability and performance under model class uncertainty.
The contributions of these papers are as follows: Paper I: Presents a min- imax optimal output prediction algorithm for linear systems with parameter uncertainty. Paper II: Proposes an algorithm to compute performance bounds for minimax adaptive estimators. Paper III: Develops a minimax suboptimal adaptive controller for scalar linear systems with noisy measurements. Paper IV: Introduces a class of nonlinear systems for which minimax dual control admits a finite-dimensional sufficient statistic, builds dynamic programming theory around this class, and designs an adaptive controller for stabilizing an integrator from absolute-value measurements. Paper V: Provides a unified framework for state-feedback and output-feedback minimax adaptive control and methods for synthesizing suboptimal controllers. Complementing these theoretical contributions are two software artifacts: one for adaptive control and the other for adaptive estimation.
The contributions apply to simple systems that represent components of larger systems, marking a step towards automating controller synthesis and maintenance for critical infrastructures. (Less)
Please use this url to cite or link to this publication:
https://lup.lub.lu.se/record/23ec674b-beec-44c3-8259-02a755cd985d
- author
- Kjellqvist, Olle LU
- supervisor
- opponent
-
- Senior Research Fellow, Dr Umenberger, Jack, University of Oxford
- organization
- alternative title
- Minimax adaptiv reglering och estimering
- publishing date
- 2024
- type
- Thesis
- publication status
- published
- subject
- pages
- 156 pages
- publisher
- Department of Automatic Control, Lund University
- defense location
- Lecture hall A, building M, Ole Römers väg 1
- defense date
- 2024-10-11 09:15:00
- ISBN
- 978-91-8104-168-2
- 978-91-8104-167-5
- project
- Scalable Control using Learning and Adaptation
- Scalable Control of Interconnected Systems
- language
- English
- LU publication?
- yes
- id
- 23ec674b-beec-44c3-8259-02a755cd985d
- date added to LUP
- 2024-08-30 14:26:59
- date last changed
- 2024-09-16 08:45:23
@phdthesis{23ec674b-beec-44c3-8259-02a755cd985d, abstract = {{This thesis presents five papers on minimax adaptive control and estimation. Minimax adaptive estimation is a framework for output prediction and state estimation that provides a priori computable performance bounds for esti- mators. Minimax adaptive controllers ensure that the closed loop has finite gain, maintaining stability and performance under model class uncertainty. <br/> The contributions of these papers are as follows: Paper I: Presents a min- imax optimal output prediction algorithm for linear systems with parameter uncertainty. Paper II: Proposes an algorithm to compute performance bounds for minimax adaptive estimators. Paper III: Develops a minimax suboptimal adaptive controller for scalar linear systems with noisy measurements. Paper IV: Introduces a class of nonlinear systems for which minimax dual control admits a finite-dimensional sufficient statistic, builds dynamic programming theory around this class, and designs an adaptive controller for stabilizing an integrator from absolute-value measurements. Paper V: Provides a unified framework for state-feedback and output-feedback minimax adaptive control and methods for synthesizing suboptimal controllers. Complementing these theoretical contributions are two software artifacts: one for adaptive control and the other for adaptive estimation.<br/> The contributions apply to simple systems that represent components of larger systems, marking a step towards automating controller synthesis and maintenance for critical infrastructures.}}, author = {{Kjellqvist, Olle}}, isbn = {{978-91-8104-168-2}}, language = {{eng}}, publisher = {{Department of Automatic Control, Lund University}}, school = {{Lund University}}, title = {{Minimax Adaptive Control and Estimation}}, url = {{https://lup.lub.lu.se/search/files/194244008/thesis.pdf}}, year = {{2024}}, }