Schaub, Philipp Leander (2024)
Redundancy Analysis and Structured Control Design for Dynamic Systems Using Set-Theoretical Methods.
Technische Universität Darmstadt
doi: 10.26083/tuprints-00028171
Dissertation, Erstveröffentlichung, Verlagsversion
Kurzbeschreibung (Abstract)
Automation advances further and further to fields in which highly complex and safety-critical system are operated autonomously. This is motivated by an enhanced quality of produced goods, an increase in profitability, or the demographic development in many industrialized countries. Systems must therefore be designed to operate autonomously, even under fault conditions. That is, faults must be diagnosed and controllers must be adapted automatically. This effort is, however, only fruitful if the underlying dynamic system is physically capable of maintaining a minimum level of operability in every possible fault scenario.
This thesis approaches this exact issue by developing a theoretic framework for analyzing dynamic systems for their redundancy properties. It is based on a newly formulated definition of redundancy merging intuitions with system theoretic terminology. Based on a reachability analysis, two types of redundantly reachable sets are introduced for redundancy analysis. Mathematical properties, applicability to system classes, and their computation are discussed in detail. A redundancy measure comprising two scalars is introduced for condensing possibly high dimensional reachability information in a manageable format. Conditions for redundant stabilizability of given reference points are formulated and verified. A respective algorithm for checking the conditions is presented.
A second contribution of the thesis is made in the field of structured control design. Imposing structural requirements on the closed-loop transfer behavior of multiple-input-multiple-output systems is necessary if certain signal paths from input to output must be decoupled from each other. Set theoretic methods are used to transform the structural design problem into equality constraints that must be satisfied by the controller and prefilter parameters of a time-invariant state feedback law. On the one hand, these constraints can be used to investigate the closed loop structure for robustness properties with respect to parameter uncertainties of the underlying plant. On the other hand, the approach enables a strictly structured method to design controllers establishing a desired closed loop structure. Combined with pole region assignment, an almost automatic structured control design is enabled.
In extension to the previous, the well-established method of linear regulator design (LQR) is investigated in the context of structured controllers. That is, the weighting matrices of the standard LQR are sought such that the resulting optimal controller fulfills the imposed equality constraints. The developed procedure makes use of results in the field of inverse optimal control and serves as an alternative for populating the degrees of freedom of structured state feedback controllers.
Typ des Eintrags: | Dissertation | ||||
---|---|---|---|---|---|
Erschienen: | 2024 | ||||
Autor(en): | Schaub, Philipp Leander | ||||
Art des Eintrags: | Erstveröffentlichung | ||||
Titel: | Redundancy Analysis and Structured Control Design for Dynamic Systems Using Set-Theoretical Methods | ||||
Sprache: | Englisch | ||||
Referenten: | Konigorski, Prof. Dr. Ulrich ; Adamy, Prof. Dr. Jürgen | ||||
Publikationsjahr: | 17 Oktober 2024 | ||||
Ort: | Darmstadt | ||||
Kollation: | XII, 135 Seiten | ||||
Datum der mündlichen Prüfung: | 24 September 2024 | ||||
DOI: | 10.26083/tuprints-00028171 | ||||
URL / URN: | https://tuprints.ulb.tu-darmstadt.de/28171 | ||||
Kurzbeschreibung (Abstract): | Automation advances further and further to fields in which highly complex and safety-critical system are operated autonomously. This is motivated by an enhanced quality of produced goods, an increase in profitability, or the demographic development in many industrialized countries. Systems must therefore be designed to operate autonomously, even under fault conditions. That is, faults must be diagnosed and controllers must be adapted automatically. This effort is, however, only fruitful if the underlying dynamic system is physically capable of maintaining a minimum level of operability in every possible fault scenario. This thesis approaches this exact issue by developing a theoretic framework for analyzing dynamic systems for their redundancy properties. It is based on a newly formulated definition of redundancy merging intuitions with system theoretic terminology. Based on a reachability analysis, two types of redundantly reachable sets are introduced for redundancy analysis. Mathematical properties, applicability to system classes, and their computation are discussed in detail. A redundancy measure comprising two scalars is introduced for condensing possibly high dimensional reachability information in a manageable format. Conditions for redundant stabilizability of given reference points are formulated and verified. A respective algorithm for checking the conditions is presented. A second contribution of the thesis is made in the field of structured control design. Imposing structural requirements on the closed-loop transfer behavior of multiple-input-multiple-output systems is necessary if certain signal paths from input to output must be decoupled from each other. Set theoretic methods are used to transform the structural design problem into equality constraints that must be satisfied by the controller and prefilter parameters of a time-invariant state feedback law. On the one hand, these constraints can be used to investigate the closed loop structure for robustness properties with respect to parameter uncertainties of the underlying plant. On the other hand, the approach enables a strictly structured method to design controllers establishing a desired closed loop structure. Combined with pole region assignment, an almost automatic structured control design is enabled. In extension to the previous, the well-established method of linear regulator design (LQR) is investigated in the context of structured controllers. That is, the weighting matrices of the standard LQR are sought such that the resulting optimal controller fulfills the imposed equality constraints. The developed procedure makes use of results in the field of inverse optimal control and serves as an alternative for populating the degrees of freedom of structured state feedback controllers. |
||||
Alternatives oder übersetztes Abstract: |
|
||||
Status: | Verlagsversion | ||||
URN: | urn:nbn:de:tuda-tuprints-281719 | ||||
Sachgruppe der Dewey Dezimalklassifikatin (DDC): | 600 Technik, Medizin, angewandte Wissenschaften > 620 Ingenieurwissenschaften und Maschinenbau | ||||
Fachbereich(e)/-gebiet(e): | 18 Fachbereich Elektrotechnik und Informationstechnik 18 Fachbereich Elektrotechnik und Informationstechnik > Institut für Automatisierungstechnik und Mechatronik 18 Fachbereich Elektrotechnik und Informationstechnik > Institut für Automatisierungstechnik und Mechatronik > Regelungstechnik und Mechatronik |
||||
Hinterlegungsdatum: | 17 Okt 2024 12:16 | ||||
Letzte Änderung: | 22 Okt 2024 13:02 | ||||
PPN: | |||||
Referenten: | Konigorski, Prof. Dr. Ulrich ; Adamy, Prof. Dr. Jürgen | ||||
Datum der mündlichen Prüfung / Verteidigung / mdl. Prüfung: | 24 September 2024 | ||||
Export: | |||||
Suche nach Titel in: | TUfind oder in Google |
Frage zum Eintrag |
Optionen (nur für Redakteure)
Redaktionelle Details anzeigen |