<div class="csl-bib-body">
<div class="csl-entry">Lukina, A. (2018). <i>Adaptive optimization framework for verification and control of cyber-physical systems</i> [Dissertation, Technische Universität Wien]. reposiTUm. https://doi.org/10.34726/hss.2018.68341</div>
</div>
-
dc.identifier.uri
https://doi.org/10.34726/hss.2018.68341
-
dc.identifier.uri
http://hdl.handle.net/20.500.12708/8606
-
dc.description.abstract
The main focus of this thesis is an optimization-based framework for control and verification of cyber-physical systems that synthesizes actions steering a given system towards a specified state. The primary motivation for the research presented in this thesis is a fascination with birds, which save energy on long-distance flights via forming a V-shape. We ask the following question: Are V-formations a result of solving an optimization problem and can this concept be utilized in cyber-physical systems, particularly in drone swarms, to increase their safety and resilience? In this thesis, we combine the state-of-the-art in statistical model checking and optimizationbased control for nonlinear systems. First, we propose a control-based evaluation of the probability of rare events that can lead to fatal accidents involving cyber-physical systems. Second, we synthesize controllers for a flock of birds modeled as a stochastic multi-agent system equipped with a highly nonlinear cost function. Further, we show that the proposed control approach is stable with respect to rare events, i.e., it leads to a successful recovery of the optimal formation from several types of malicious attacks. Finally, we investigate the ways to efficiently distribute control among the agents. We demonstrate that our framework can be applied to any system modeled as a controllable Markov decision process with a cost (reward) function. A key feature of the procedure we propose is its automatic adaptation to the convergence performance of optimization towards a given global objective. Combining model-predictive control and ideas from sequential Monte-Carlo methods, we introduce a performance-based adaptive horizon and implicitly build a Lyapunov function that guarantees convergence. We use statistical model-checking to verify the algorithm and assess its reliability.
en
dc.language
English
-
dc.language.iso
en
-
dc.rights.uri
http://rightsstatements.org/vocab/InC/1.0/
-
dc.subject
cyber physical systems
en
dc.subject
control synthesis
en
dc.subject
statistical model checking
en
dc.subject
Markov decision processes
en
dc.subject
model predictive control
en
dc.subject
flocking
en
dc.subject
adaptive optimization
en
dc.subject
drones
en
dc.subject
V formation
en
dc.subject
sequential Monte-Carlo methods
en
dc.title
Adaptive optimization framework for verification and control of cyber-physical systems
en
dc.type
Thesis
en
dc.type
Hochschulschrift
de
dc.rights.license
In Copyright
en
dc.rights.license
Urheberrechtsschutz
de
dc.identifier.doi
10.34726/hss.2018.68341
-
dc.contributor.affiliation
TU Wien, Österreich
-
dc.rights.holder
Anna Lukina
-
dc.publisher.place
Wien
-
tuw.version
vor
-
tuw.thesisinformation
Technische Universität Wien
-
tuw.publication.orgunit
E191 - Institut für Computer Engineering
-
dc.type.qualificationlevel
Doctoral
-
dc.identifier.libraryid
AC15406841
-
dc.description.numberOfPages
110
-
dc.identifier.urn
urn:nbn:at:at-ubtuw:1-127016
-
dc.thesistype
Dissertation
de
dc.thesistype
Dissertation
en
dc.rights.identifier
In Copyright
en
dc.rights.identifier
Urheberrechtsschutz
de
tuw.advisor.staffStatus
staff
-
tuw.advisor.orcid
0000-0001-5715-2142
-
item.openairetype
doctoral thesis
-
item.fulltext
with Fulltext
-
item.cerifentitytype
Publications
-
item.openaccessfulltext
Open Access
-
item.mimetype
application/pdf
-
item.languageiso639-1
en
-
item.openairecristype
http://purl.org/coar/resource_type/c_db06
-
item.grantfulltext
open
-
crisitem.author.dept
E191-01 - Forschungsbereich Cyber-Physical Systems