2022
|
Mostacciuolo, Elisa; Trenn, Stephan; Vasca, Francesco Averaging for switched impulsive systems with pulse width modulation Unpublished 2022, (submitted). @unpublished{MostTren22ppb,
title = {Averaging for switched impulsive systems with pulse width modulation},
author = {Elisa Mostacciuolo and Stephan Trenn and Francesco Vasca},
url = {https://stephantrenn.net/wp-content/uploads/2022/11/Preprint-MTV221128.pdf, Preprint},
year = {2022},
date = {2022-11-28},
urldate = {2022-11-28},
abstract = {Linear switched impulsive systems (SIS) are characterized by ordinary differential equations as modes dynamics and state jumps at the switching time instants. The presence of possible jumps in the state makes nontrivial the application of classical averaging techniques. In this paper we consider SIS with pulse width modulation (PWM) and we propose an averaged model whose solution approximates the moving average of the SIS solution with an error which decreases with the multiple of the switching period and by decreasing the PWM period. The averaging result requires milder assumptions on the system matrices with respect to those needed by the previous averaging techniques for SIS. The interest of the proposed model is strengthened by the fact that it reduces to the classical averaged model for PWM systems when there are no jumps in the state. The theoretical results are verified through numerical results obtained by considering a switched capacitor electrical circuit.},
note = {submitted},
keywords = {application, averaging, DAEs, LMIs, switched-DAEs, switched-systems},
pubstate = {published},
tppubtype = {unpublished}
}
Linear switched impulsive systems (SIS) are characterized by ordinary differential equations as modes dynamics and state jumps at the switching time instants. The presence of possible jumps in the state makes nontrivial the application of classical averaging techniques. In this paper we consider SIS with pulse width modulation (PWM) and we propose an averaged model whose solution approximates the moving average of the SIS solution with an error which decreases with the multiple of the switching period and by decreasing the PWM period. The averaging result requires milder assumptions on the system matrices with respect to those needed by the previous averaging techniques for SIS. The interest of the proposed model is strengthened by the fact that it reduces to the classical averaged model for PWM systems when there are no jumps in the state. The theoretical results are verified through numerical results obtained by considering a switched capacitor electrical circuit. |
2020
|
Hu, Jiaming; Trenn, Stephan Sliding mode observer based hysteresis compensation control for piezoelectric stacks Miscellaneous Book of Abstracts - 39th Benelux Meeting on Systems and Control, 2020. @misc{HuTren20m,
title = {Sliding mode observer based hysteresis compensation control for piezoelectric stacks},
author = {Jiaming Hu and Stephan Trenn},
editor = {Raffaella Carloni and Bayu Jayawardhana and Erjen Lefeber},
url = {https://www.beneluxmeeting.nl/2020/uploads/papers/boa.pdf, Book of Abstracts
https://stephantrenn.net/wp-content/uploads/2021/03/HuTren20.pdf, Extended Abstract},
year = {2020},
date = {2020-03-12},
howpublished = {Book of Abstracts - 39th Benelux Meeting on Systems and Control},
keywords = {application, nonlinear},
pubstate = {published},
tppubtype = {misc}
}
|
Boon, Marko; van den Bosch, Mark; Breeuwsma, Paul; Bucchianico, Alessandro Di; Emampour, Mona; van Ginkel, Bart; Hepkema, Tjebbe; Holzinger, Philipp; Timmerman, Rik; Trenn, Stephan Event Driven Model with an Objective to Control Traffic Lights in the Netherlands Miscellaneous Scientific Proceedings 157th European Study Group with Industry (SWI 2020), 2020. @misc{BoonBosc20m,
title = {Event Driven Model with an Objective to Control Traffic Lights in the Netherlands},
author = {Marko Boon and Mark van den Bosch and Paul Breeuwsma and Alessandro Di Bucchianico and Mona Emampour and Bart van Ginkel and Tjebbe Hepkema and Philipp Holzinger and Rik Timmerman and Stephan Trenn},
editor = {Bram van den Broek},
url = {https://stephantrenn.net/wp-content/uploads/2023/02/BoonBosc20m.pdf, Report
https://www.swi-wiskunde.nl/swi2020/wp-content/uploads/sites/26/2020/10/swi2020scproceedings.pdf, Proceedings SWI 2020},
year = {2020},
date = {2020-03-01},
urldate = {2020-03-01},
abstract = {The study group participants of SWI 2020 with regard to the challenge proposed by the company Sweco were tasked to initiate a discrete-event dynamic model into Smart Traffic. Smart Traffic is cloud driven software developed by Sweco, implementing real-time predictive traf- fic signal control. Currently, the microscopic traffic simulator SUMO is being used within Smart Traffic to predict the traffic pattern for the short-term future, with the purpose of optimising traffic signal settings. However, in practice, microscopic traffic simulators appear to be too slow and hence infeasible considering its application. We employ discrete-event simulations as a tool to predict the future traffic state both efficiently and effectively, even though those type of simulations are usually employed in a different context.
We were in particular advised to focus on devising an event-driven model for a single, general intersection. This enabled us to create a thorough mathematical basic model. We are able to study various performance characteristics of the traffic light, such as the total delay or to the total squared delay. Accompanied with the mathematical basic model, we deliver in correspondence a fully functioning program written in Python. Our article includes a detailed yet relatively simple example based on this program. This example additionally demonstrates the difference in an optimal outcome when using the total delay or the total squared delay. Ultimately, we note that our model is easily extendable and several feasible extensions are proposed in this article.},
howpublished = {Scientific Proceedings 157th European Study Group with Industry (SWI 2020)},
keywords = {application},
pubstate = {published},
tppubtype = {misc}
}
The study group participants of SWI 2020 with regard to the challenge proposed by the company Sweco were tasked to initiate a discrete-event dynamic model into Smart Traffic. Smart Traffic is cloud driven software developed by Sweco, implementing real-time predictive traf- fic signal control. Currently, the microscopic traffic simulator SUMO is being used within Smart Traffic to predict the traffic pattern for the short-term future, with the purpose of optimising traffic signal settings. However, in practice, microscopic traffic simulators appear to be too slow and hence infeasible considering its application. We employ discrete-event simulations as a tool to predict the future traffic state both efficiently and effectively, even though those type of simulations are usually employed in a different context.
We were in particular advised to focus on devising an event-driven model for a single, general intersection. This enabled us to create a thorough mathematical basic model. We are able to study various performance characteristics of the traffic light, such as the total delay or to the total squared delay. Accompanied with the mathematical basic model, we deliver in correspondence a fully functioning program written in Python. Our article includes a detailed yet relatively simple example based on this program. This example additionally demonstrates the difference in an optimal outcome when using the total delay or the total squared delay. Ultimately, we note that our model is easily extendable and several feasible extensions are proposed in this article. |
2018
|
Gross, Tjorben B.; Trenn, Stephan; Wirsen, Andreas Switch induced instabilities for stable power system DAE models Inproceedings In: IFAC-PapersOnLine, pp. 127-132, 2018, (Proc. IFAC Conf. Analysis Design Hybrid Systems (ADHS 2018)). @inproceedings{GrosTren18,
title = {Switch induced instabilities for stable power system DAE models},
author = {Tjorben B. Gross and Stephan Trenn and Andreas Wirsen},
url = {https://stephantrenn.net/wp-content/uploads/2018/04/Preprint-GTW180413.pdf, Preprint},
doi = {10.1016/j.ifacol.2018.08.022},
year = {2018},
date = {2018-07-11},
booktitle = {IFAC-PapersOnLine},
journal = {IFAC-PapersOnLine},
volume = {51},
number = {16},
pages = {127-132},
abstract = {It is well known that for switched systems the overall dynamics can be unstable despite stability of all individual modes. We show that this phenoma can indeed occur for a linearized DAE model of power grids. By making certain topological assumptions on the power grid, we can ensure stability under arbitrary switching.},
note = {Proc. IFAC Conf. Analysis Design Hybrid Systems (ADHS 2018)},
keywords = {application, stability, switched-DAEs, switched-systems},
pubstate = {published},
tppubtype = {inproceedings}
}
It is well known that for switched systems the overall dynamics can be unstable despite stability of all individual modes. We show that this phenoma can indeed occur for a linearized DAE model of power grids. By making certain topological assumptions on the power grid, we can ensure stability under arbitrary switching. |
Kausar, Rukhsana; Trenn, Stephan Water hammer modeling for water networks via hyperbolic PDEs and switched DAEs Inproceedings In: Klingenberg, Christian; Westdickenberg, Michael (Ed.): Theory, Numerics and Applications of Hyperbolic Problems II, pp. 123-135, Springer, Cham, 2018, ISBN: 978-3-319-91548-7, (Presented at XVI International Conference on Hyperbolic Problems (HYP2016), Aachen). @inproceedings{KausTren18,
title = {Water hammer modeling for water networks via hyperbolic PDEs and switched DAEs},
author = {Rukhsana Kausar and Stephan Trenn},
editor = {Christian Klingenberg and Michael Westdickenberg},
url = {https://stephantrenn.net/wp-content/uploads/2017/09/Preprint-KT170418.pdf, Preprint},
doi = {10.1007/978-3-319-91548-7_9},
isbn = {978-3-319-91548-7},
year = {2018},
date = {2018-06-27},
urldate = {2018-06-27},
booktitle = {Theory, Numerics and Applications of Hyperbolic Problems II},
pages = {123-135},
publisher = {Springer},
address = {Cham},
abstract = {In water distribution network instantaneous changes in valve and pump settings introduce jumps and sometimes impulses. In particular, a particular impulsive phenomenon which occurs due to sudden closing of valve is the so called water hammer. It is classically modeled as a system of hyperbolic partial differential equations (PDEs). We observed that under some suitable assumptions the PDEs usually used to describe water flows can be simplified to differential algebraic equations (DAEs). The idea is to model water hammer phenomenon in the switched DAEs framework due to its special feature of studying such impulsive effects. To compare these two modeling techniques, a system of hyperbolic PDE model and the switched DAE model for a simple set up consisting of two reservoirs, six pipes and three valve is presented. The aim of this contribution is to present results of both models as motivation for the claim that a switched DAE modeling framework is suitable for describing a water hammer.},
note = {Presented at XVI International Conference on Hyperbolic Problems (HYP2016), Aachen},
keywords = {application, DAEs, nonlinear, PDEs, piecewise-smooth-distributions, solution-theory, switched-DAEs, switched-systems},
pubstate = {published},
tppubtype = {inproceedings}
}
In water distribution network instantaneous changes in valve and pump settings introduce jumps and sometimes impulses. In particular, a particular impulsive phenomenon which occurs due to sudden closing of valve is the so called water hammer. It is classically modeled as a system of hyperbolic partial differential equations (PDEs). We observed that under some suitable assumptions the PDEs usually used to describe water flows can be simplified to differential algebraic equations (DAEs). The idea is to model water hammer phenomenon in the switched DAEs framework due to its special feature of studying such impulsive effects. To compare these two modeling techniques, a system of hyperbolic PDE model and the switched DAE model for a simple set up consisting of two reservoirs, six pipes and three valve is presented. The aim of this contribution is to present results of both models as motivation for the claim that a switched DAE modeling framework is suitable for describing a water hammer. |
2017
|
Kausar, Rukhsana; Trenn, Stephan Impulses in structured nonlinear switched DAEs Inproceedings In: Proc. 56th IEEE Conf. Decis. Control, pp. 3181 - 3186, Melbourne, Australia, 2017. @inproceedings{KausTren17b,
title = {Impulses in structured nonlinear switched DAEs},
author = {Rukhsana Kausar and Stephan Trenn},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-KT170920.pdf, Preprint},
doi = {10.1109/CDC.2017.8264125},
year = {2017},
date = {2017-12-14},
booktitle = {Proc. 56th IEEE Conf. Decis. Control},
pages = {3181 - 3186},
address = {Melbourne, Australia},
abstract = { Switched nonlinear differential algebraic equations (DAEs) occur in mathematical modeling of sudden transients in various physical phenomenons. Hence, it is important to investigate them with respect to the nature of their solutions. The few existing solvability results for switched nonlinear DAEs exclude Dirac impulses by definition; however, in many cases this is too restrictive. For example, in water distribution networks the water hammer effect can only be studied when allowing Dirac impulses in a nonlinear switched DAE description. We investigate existence and uniqueness of solutions with impulses for a general class of nonlinear switched DAEs, where we exploit a certain sparse structure of the nonlinearity.},
keywords = {application, CDC, DAEs, nonlinear, piecewise-smooth-distributions, solution-theory, switched-DAEs, switched-systems},
pubstate = {published},
tppubtype = {inproceedings}
}
Switched nonlinear differential algebraic equations (DAEs) occur in mathematical modeling of sudden transients in various physical phenomenons. Hence, it is important to investigate them with respect to the nature of their solutions. The few existing solvability results for switched nonlinear DAEs exclude Dirac impulses by definition; however, in many cases this is too restrictive. For example, in water distribution networks the water hammer effect can only be studied when allowing Dirac impulses in a nonlinear switched DAE description. We investigate existence and uniqueness of solutions with impulses for a general class of nonlinear switched DAEs, where we exploit a certain sparse structure of the nonlinearity. |
Küsters, Ferdinand; Patil, Deepak; Tesi, Pietro; Trenn, Stephan Indiscernible topological variations in DAE networks with applications to power grids Inproceedings In: Proc. 20th IFAC World Congress 2017, pp. 7333 - 7338, Toulouse, France, 2017, ISSN: 2405-8963. @inproceedings{KustPati17a,
title = {Indiscernible topological variations in DAE networks with applications to power grids},
author = {Ferdinand Küsters and Deepak Patil and Pietro Tesi and Stephan Trenn},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-KPTT170320.pdf, Preprint},
doi = {10.1016/j.ifacol.2017.08.1478},
issn = {2405-8963},
year = {2017},
date = {2017-03-24},
booktitle = {Proc. 20th IFAC World Congress 2017},
journal = {IFAC-PapersOnLine},
volume = {50},
number = {1},
pages = {7333 - 7338},
address = {Toulouse, France},
abstract = {The ability to detect topology variations in dynamical networks defined by differential algebraic equations (DAEs) is considered. We characterize the existence of initial states, for which topological changes are indiscernible. A key feature of our characterization is the ability to verify indiscernibility just in terms of the nominal topology. We apply the results to a power grid model and also discuss the relationship to recent mode-detection results for switched DAEs.},
keywords = {application, DAEs, networks, observability},
pubstate = {published},
tppubtype = {inproceedings}
}
The ability to detect topology variations in dynamical networks defined by differential algebraic equations (DAEs) is considered. We characterize the existence of initial states, for which topological changes are indiscernible. A key feature of our characterization is the ability to verify indiscernibility just in terms of the nominal topology. We apply the results to a power grid model and also discuss the relationship to recent mode-detection results for switched DAEs. |
Kall, Jochen; Kausar, Rukhsana; Trenn, Stephan Modeling water hammers via PDEs and switched DAEs with numerical justification Inproceedings In: Proc. 20th IFAC World Congress 2017, pp. 5349 - 5354, Toulouse, France, 2017, ISSN: 2405-8963. @inproceedings{KallKaus17,
title = {Modeling water hammers via PDEs and switched DAEs with numerical justification},
author = {Jochen Kall and Rukhsana Kausar and Stephan Trenn},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-KKT170324.pdf, Preprint},
doi = {10.1016/j.ifacol.2017.08.927},
issn = {2405-8963},
year = {2017},
date = {2017-03-23},
booktitle = {Proc. 20th IFAC World Congress 2017},
journal = {IFAC-PapersOnLine},
volume = {50},
number = {1},
pages = {5349 - 5354},
address = {Toulouse, France},
abstract = {In water distribution networks instantaneous changes in valve and pump settings may introduces jumps and peaks in the pressure. In particular, a well known phenomenon in response to the sudden closing of a valve is the so called water hammer, which (if not taken into account properly) may destroy parts of the water network. It is classically modeled as a system of hyperbolic partial differential equations (PDEs). After discussing this PDE model we propose a simplified model using switched differential-algebraic equations (DAEs). Switched DAEs are known to be able to produce infinite peaks in response to sudden structural changes. These peaks (in the mathematical form of Dirac impulses) can easily be predicted and may allow for a simpler analysis of complex water networks in the future. As a first step toward that goal, we verify the novel modeling approach by comparing these two modeling techniques numerically for a simple set up consisting of two reservoirs, a pipe and a valve.},
keywords = {application, DAEs, nonlinear, PDEs, solution-theory, switched-DAEs, switched-systems},
pubstate = {published},
tppubtype = {inproceedings}
}
In water distribution networks instantaneous changes in valve and pump settings may introduces jumps and peaks in the pressure. In particular, a well known phenomenon in response to the sudden closing of a valve is the so called water hammer, which (if not taken into account properly) may destroy parts of the water network. It is classically modeled as a system of hyperbolic partial differential equations (PDEs). After discussing this PDE model we propose a simplified model using switched differential-algebraic equations (DAEs). Switched DAEs are known to be able to produce infinite peaks in response to sudden structural changes. These peaks (in the mathematical form of Dirac impulses) can easily be predicted and may allow for a simpler analysis of complex water networks in the future. As a first step toward that goal, we verify the novel modeling approach by comparing these two modeling techniques numerically for a simple set up consisting of two reservoirs, a pipe and a valve. |
2016
|
Gross, Tjorben B.; Trenn, Stephan; Wirsen, Andreas Solvability and stability of a power system DAE model Journal Article In: Syst. Control Lett., vol. 97, pp. 12–17, 2016. @article{GrosTren16,
title = {Solvability and stability of a power system DAE model},
author = {Tjorben B. Gross and Stephan Trenn and Andreas Wirsen},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-GTW160816.pdf, Preprint},
doi = {10.1016/j.sysconle.2016.08.003},
year = {2016},
date = {2016-11-01},
journal = {Syst. Control Lett.},
volume = {97},
pages = {12--17},
abstract = {The dynamic model of a power system is the combination of the power flow equations and the dynamic description of the generators (the swing equations) resulting in a differential–algebraic equation (DAE). For general DAEs solvability is not guaranteed in general, in the linear case the coefficient matrices have to satisfy a certain regularity condition. We derive a solvability characterization for the linearized power system DAE solely in terms of the network topology. As an extension to previous result we allow for higher order generator dynamics. Furthermore, we show that any solvable power system DAE is automatically of index one, which means that it is also numerically well posed. Finally, we show that any solvable power system DAE is stable but not asymptotically stable.},
keywords = {application, DAEs, Lyapunov, networks, solution-theory, stability},
pubstate = {published},
tppubtype = {article}
}
The dynamic model of a power system is the combination of the power flow equations and the dynamic description of the generators (the swing equations) resulting in a differential–algebraic equation (DAE). For general DAEs solvability is not guaranteed in general, in the linear case the coefficient matrices have to satisfy a certain regularity condition. We derive a solvability characterization for the linearized power system DAE solely in terms of the network topology. As an extension to previous result we allow for higher order generator dynamics. Furthermore, we show that any solvable power system DAE is automatically of index one, which means that it is also numerically well posed. Finally, we show that any solvable power system DAE is stable but not asymptotically stable. |
2015
|
Mostacciuolo, Elisa; Trenn, Stephan; Vasca, Francesco Averaging for non-homogeneous switched DAEs Inproceedings In: Proc. 54th IEEE Conf. Decis. Control, Osaka, Japan, pp. 2951–2956, 2015. @inproceedings{MostTren15b,
title = {Averaging for non-homogeneous switched DAEs},
author = {Elisa Mostacciuolo and Stephan Trenn and Francesco Vasca},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-MTV150901.pdf, Preprint},
doi = {10.1109/CDC.2015.7402665},
year = {2015},
date = {2015-12-02},
booktitle = {Proc. 54th IEEE Conf. Decis. Control, Osaka, Japan},
pages = {2951--2956},
abstract = {Averaging is widely used for approximating the dynamics of switched systems. The validity of an averaged model typically depends on the switching frequency and on some technicalities regarding the switched system structure. For homogeneous linear switched differential algebraic equations it is known that an averaged model can be obtained. In this paper an averaging result for non-homogeneous switched systems is presented. A switched electrical circuit illustrates the practical interest of the result.},
keywords = {application, averaging, CDC, DAEs, switched-DAEs, switched-systems},
pubstate = {published},
tppubtype = {inproceedings}
}
Averaging is widely used for approximating the dynamics of switched systems. The validity of an averaged model typically depends on the switching frequency and on some technicalities regarding the switched system structure. For homogeneous linear switched differential algebraic equations it is known that an averaged model can be obtained. In this paper an averaging result for non-homogeneous switched systems is presented. A switched electrical circuit illustrates the practical interest of the result. |
2014
|
Gross, Tjorben B.; Trenn, Stephan; Wirsen, Andreas Topological solvability and index characterizations for a common DAE power system model Inproceedings In: Proc. 2014 IEEE Conf. Control Applications (CCA), pp. 9–14, IEEE 2014. @inproceedings{GrosTren14,
title = {Topological solvability and index characterizations for a common DAE power system model},
author = {Tjorben B. Gross and Stephan Trenn and Andreas Wirsen},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-GTW140904.pdf, Preprint},
doi = {10.1109/CCA.2014.6981321},
year = {2014},
date = {2014-10-10},
booktitle = {Proc. 2014 IEEE Conf. Control Applications (CCA)},
pages = {9--14},
organization = {IEEE},
abstract = {For the widely-used power system model consisting of the generator swing equations and the power flow equations resulting in a system of differential algebraic equations (DAEs), we introduce a sufficient and necessary solvability condition for the linearized model. This condition is based on the topological structure of the power system. Furthermore we show sufficient conditions for the linearized DAE-system and a nonlinear version of the model to have differentiation index equal to one.},
keywords = {application, DAEs, networks, nonlinear, solution-theory},
pubstate = {published},
tppubtype = {inproceedings}
}
For the widely-used power system model consisting of the generator swing equations and the power flow equations resulting in a system of differential algebraic equations (DAEs), we introduce a sufficient and necessary solvability condition for the linearized model. This condition is based on the topological structure of the power system. Furthermore we show sufficient conditions for the linearized DAE-system and a nonlinear version of the model to have differentiation index equal to one. |
2013
|
Liberzon, Daniel; Trenn, Stephan The bang-bang funnel controller: time delays and case study Inproceedings In: Proc. 12th European Control Conf. (ECC) 2013, Zurich, Switzerland, pp. 1669–1674, 2013. @inproceedings{LibeTren13a,
title = {The bang-bang funnel controller: time delays and case study},
author = {Daniel Liberzon and Stephan Trenn},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-LT130320.pdf, Preprint
http://ieeexplore.ieee.org/document/6669120, IEEE Xplore Article Number 6669120},
year = {2013},
date = {2013-07-01},
booktitle = {Proc. 12th European Control Conf. (ECC) 2013, Zurich, Switzerland},
pages = {1669--1674},
abstract = {We investigate the recently introduced bang-bang funnel controller with respect to its robustness to time delays. We present slightly modified feasibility conditions and prove that the bang-bang funnel controller applied to a relative-degree-two nonlinear system can tolerate sufficiently small time delays. A second contribution of this paper is an extensive case study, based on a model of a real experimental setup, where implementation issues such as the necessary sampling time and the conservativeness of the feasibility assumptions are explicitly considered.},
keywords = {application, funnel-control, input-constraints, nonlinear, relative-degree},
pubstate = {published},
tppubtype = {inproceedings}
}
We investigate the recently introduced bang-bang funnel controller with respect to its robustness to time delays. We present slightly modified feasibility conditions and prove that the bang-bang funnel controller applied to a relative-degree-two nonlinear system can tolerate sufficiently small time delays. A second contribution of this paper is an extensive case study, based on a model of a real experimental setup, where implementation issues such as the necessary sampling time and the conservativeness of the feasibility assumptions are explicitly considered. |
Hackl, Christoph M.; Hopfe, Norman; Ilchmann, Achim; Mueller, Markus; Trenn, Stephan Funnel control for systems with relative degree two Journal Article In: SIAM J. Control Optim., vol. 51, no. 2, pp. 965–995, 2013. @article{HackHopf13,
title = {Funnel control for systems with relative degree two},
author = {Christoph M. Hackl and Norman Hopfe and Achim Ilchmann and Markus Mueller and Stephan Trenn},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/HackHopf13.pdf, Paper},
doi = {10.1137/100799903 },
year = {2013},
date = {2013-03-19},
journal = {SIAM J. Control Optim.},
volume = {51},
number = {2},
pages = {965--995},
abstract = {Tracking of reference signals y_ref(.) by the output y(.) of linear (as well as a considerably large class of nonlinear) single-input, single-output systems is considered. The system is assumed to have strict relative degree two with (weakly) stable zero dynamics. The control objective is tracking of the error e=y-y_ref and its derivative e' within two prespecified performance funnels, respectively. This is achieved by the so-called funnel controller u(t) = -k_0(t)^2 e(t) - k_1(t) e'(t), where the simple proportional error feedback has gain functions k_0 and k_1 designed in such a way to preclude contact of e and e' with the funnel boundaries, respectively. The funnel controller also ensures boundedness of all signals. We also show that the same funnel controller (i) is applicable to relative degree one systems, (ii) allows for input constraints provided a feasibility condition (formulated in terms of the system data, the saturation bounds, the funnel data, bounds on the reference signal, and the initial state) holds, (iii) is robust in terms of the gap metric: if a system is sufficiently close to a system with relative degree two, stable zero dynamics, and positive high-frequency gain, but does not necessarily have these properties, then for small initial values the funnel controller also achieves the control objective. Finally, we illustrate the theoretical results by experimental results: the funnel controller is applied to a rotatory mechanical system for position control.},
keywords = {application, funnel-control, input-constraints, nonlinear, relative-degree},
pubstate = {published},
tppubtype = {article}
}
Tracking of reference signals y_ref(.) by the output y(.) of linear (as well as a considerably large class of nonlinear) single-input, single-output systems is considered. The system is assumed to have strict relative degree two with (weakly) stable zero dynamics. The control objective is tracking of the error e=y-y_ref and its derivative e' within two prespecified performance funnels, respectively. This is achieved by the so-called funnel controller u(t) = -k_0(t)^2 e(t) - k_1(t) e'(t), where the simple proportional error feedback has gain functions k_0 and k_1 designed in such a way to preclude contact of e and e' with the funnel boundaries, respectively. The funnel controller also ensures boundedness of all signals. We also show that the same funnel controller (i) is applicable to relative degree one systems, (ii) allows for input constraints provided a feasibility condition (formulated in terms of the system data, the saturation bounds, the funnel data, bounds on the reference signal, and the initial state) holds, (iii) is robust in terms of the gap metric: if a system is sufficiently close to a system with relative degree two, stable zero dynamics, and positive high-frequency gain, but does not necessarily have these properties, then for small initial values the funnel controller also achieves the control objective. Finally, we illustrate the theoretical results by experimental results: the funnel controller is applied to a rotatory mechanical system for position control. |
2012
|
Hackl, Christoph M.; Trenn, Stephan The bang-bang funnel controller: An experimental verification Inproceedings In: PAMM - Proc. Appl. Math. Mech., pp. 735–736, GAMM Annual Meeting 2012, Darmstadt Wiley-VCH Verlag GmbH, Weinheim, 2012. @inproceedings{HackTren12,
title = {The bang-bang funnel controller: An experimental verification},
author = {Christoph M. Hackl and Stephan Trenn},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-HT120427.pdf, Preprint},
doi = {10.1002/pamm.201210356},
year = {2012},
date = {2012-03-01},
booktitle = {PAMM - Proc. Appl. Math. Mech.},
volume = {12},
number = {1},
pages = {735--736},
publisher = {Wiley-VCH Verlag GmbH},
address = {Weinheim},
organization = {GAMM Annual Meeting 2012, Darmstadt},
abstract = {We adjust the newly developed bang-bang funnel controller such that it is more applicable for real world scenarios. The main idea is to introduce a third “neutral” input value to account for the situation when the error is already small enough and no control action is necessary. We present experimental results to illustrate the effectiveness of our new approach in the case of position control of an electrical drive.},
keywords = {application, funnel-control, input-constraints, nonlinear, relative-degree},
pubstate = {published},
tppubtype = {inproceedings}
}
We adjust the newly developed bang-bang funnel controller such that it is more applicable for real world scenarios. The main idea is to introduce a third “neutral” input value to account for the situation when the error is already small enough and no control action is necessary. We present experimental results to illustrate the effectiveness of our new approach in the case of position control of an electrical drive. |
2010
|
Domínguez-García, Alejandro D.; Trenn, Stephan Detection of impulsive effects in switched DAEs with applications to power electronics reliability analysis Inproceedings In: Proc. 49th IEEE Conf. Decis. Control, Atlanta, USA, pp. 5662–5667, 2010. @inproceedings{DomiTren10,
title = {Detection of impulsive effects in switched DAEs with applications to power electronics reliability analysis},
author = {Alejandro D. Domínguez-García and Stephan Trenn},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-DT100810.pdf, Preprint},
doi = {10.1109/CDC.2010.5717011},
year = {2010},
date = {2010-12-17},
booktitle = {Proc. 49th IEEE Conf. Decis. Control, Atlanta, USA},
pages = {5662--5667},
abstract = {This paper presents an analytical framework for detecting the presence of jumps and impulses in the solutions of switched differential algebraic equations (switched DAEs). The framework can be applied in the early design stage of fault-tolerant power electronics systems to identify design flaws that could jeopardize its reliability. The system is described by a switched differential algebraic equation, accounting for both fault-free system configurations and the configurations that arise after component faults, where each configuration p is defined by a pair of matrices (Ep;Ap). For each configuration p, the so called consistency projector is obtained from the pair (Ep;Ap). Based on the consistency projectors of all possible configurations, conditions for impulse-free and jump-free solutions of the switched DAE are established. A case-study of a dual redundant buck converter is presented to illustrate the framework.},
keywords = {application, CDC, DAEs, piecewise-smooth-distributions, switched-DAEs, switched-systems},
pubstate = {published},
tppubtype = {inproceedings}
}
This paper presents an analytical framework for detecting the presence of jumps and impulses in the solutions of switched differential algebraic equations (switched DAEs). The framework can be applied in the early design stage of fault-tolerant power electronics systems to identify design flaws that could jeopardize its reliability. The system is described by a switched differential algebraic equation, accounting for both fault-free system configurations and the configurations that arise after component faults, where each configuration p is defined by a pair of matrices (Ep;Ap). For each configuration p, the so called consistency projector is obtained from the pair (Ep;Ap). Based on the consistency projectors of all possible configurations, conditions for impulse-free and jump-free solutions of the switched DAE are established. A case-study of a dual redundant buck converter is presented to illustrate the framework. |
2006
|
Mandaloju, Nagendra P.; Trenn, Stephan Analogue Implementation of the funnel controller Inproceedings In: PAMM - Proc. Appl. Math. Mech., pp. 823–824, WILEY-VCH Verlag, 2006, ISSN: 1617-7061. @inproceedings{MandTren06,
title = {Analogue Implementation of the funnel controller},
author = {Nagendra P. Mandaloju and Stephan Trenn},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-MT060428.pdf, Preprint},
doi = {10.1002/pamm.200610391},
issn = {1617-7061},
year = {2006},
date = {2006-05-01},
booktitle = {PAMM - Proc. Appl. Math. Mech.},
volume = {6},
number = {1},
pages = {823--824},
publisher = {WILEY-VCH Verlag},
abstract = {In many tracking control problems, pre-specified bounds for the evolution of the tracking error should be met. The ‘funnel controller’ addresses this requirement and guarantees transient performance for a fairly large class of systems. In addition, only structural assumptions on the underlying system are made; the exact knowledge of the system parameters is not required. This is in contrast to most classical controllers where only asymptotic behaviour can be guaranteed and the system parameters must be known or estimated. Until now, the funnel controller was only studied theoretically. We will present the results of an analogue implementation of the funnel controller. The results show that the funnel controller works well in reality, i.e. it guarantees the pre-specified error bounds. The implementation is an analogue circuit composed of standard devices and is therefore suitable for a broad range of applications.},
keywords = {application, funnel-control, nonlinear},
pubstate = {published},
tppubtype = {inproceedings}
}
In many tracking control problems, pre-specified bounds for the evolution of the tracking error should be met. The ‘funnel controller’ addresses this requirement and guarantees transient performance for a fairly large class of systems. In addition, only structural assumptions on the underlying system are made; the exact knowledge of the system parameters is not required. This is in contrast to most classical controllers where only asymptotic behaviour can be guaranteed and the system parameters must be known or estimated. Until now, the funnel controller was only studied theoretically. We will present the results of an analogue implementation of the funnel controller. The results show that the funnel controller works well in reality, i.e. it guarantees the pre-specified error bounds. The implementation is an analogue circuit composed of standard devices and is therefore suitable for a broad range of applications. |
Ilchmann, Achim; Sawodny, Oliver; Trenn, Stephan Pneumatic cylinders: modelling and feedback force-control Journal Article In: Int. J. Control, vol. 79, no. 6, pp. 650–661, 2006. @article{IlchSawo06,
title = {Pneumatic cylinders: modelling and feedback force-control},
author = {Achim Ilchmann and Oliver Sawodny and Stephan Trenn},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-IST050502.pdf, Preprint},
doi = {10.1080/00207170600645875},
year = {2006},
date = {2006-01-01},
journal = {Int. J. Control},
volume = {79},
number = {6},
pages = {650--661},
abstract = {In this paper, we model, analyse, and control an experimental set-up of a servo pneumatic cylinder. The dynamic behaviour of pneumatic actuator systems is dominant by non-linear functions. First, a mathematical model for the pneumatic system is derived. Secondly, we investigate the mathematical properties of this model and show boundedness and positiveness of certain variables. Thirdly, we prove that a proportional output feedback controller with saturation achieves practical tracking a wide class of reference trajectories. We verify the theoretical results and the effectiveness of the control by experiments.},
keywords = {application, input-constraints},
pubstate = {published},
tppubtype = {article}
}
In this paper, we model, analyse, and control an experimental set-up of a servo pneumatic cylinder. The dynamic behaviour of pneumatic actuator systems is dominant by non-linear functions. First, a mathematical model for the pneumatic system is derived. Secondly, we investigate the mathematical properties of this model and show boundedness and positiveness of certain variables. Thirdly, we prove that a proportional output feedback controller with saturation achieves practical tracking a wide class of reference trajectories. We verify the theoretical results and the effectiveness of the control by experiments. |
2004
|
Ilchmann, Achim; Trenn, Stephan Input constrained funnel control with applications to chemical reactor models Journal Article In: Syst. Control Lett., vol. 53, no. 5, pp. 361–375, 2004. @article{IlchTren04,
title = {Input constrained funnel control with applications to chemical reactor models},
author = {Achim Ilchmann and Stephan Trenn},
url = {http://stephantrenn.net/wp-content/uploads/2017/09/Preprint-IT040715.pdf, Preprint},
doi = {10.1016/j.sysconle.2004.05.014},
year = {2004},
date = {2004-01-01},
journal = {Syst. Control Lett.},
volume = {53},
number = {5},
pages = {361--375},
publisher = {Elsevier},
abstract = {Error feedback control is considered for a class of exothermic chemical reactor models. The control objective is that the temperature T evolves within a prespecified performance envelope or ``funnel'' around the set point temperature T*. A simple error feedback control with input constraints of the form u(t)=sat(-k(t)[T(t)-T*] + u*), u* an offset, is introduced which achieves the objective in the presence of disturbances corrupting the measurement. The gain k(t) is a function of the error e(t)=T(t)-T* and its distance to the funnel boundary. The input constraints have to satisfy certain feasibility assumptions in terms of the model data and the operating point T*.},
keywords = {application, funnel-control, input-constraints},
pubstate = {published},
tppubtype = {article}
}
Error feedback control is considered for a class of exothermic chemical reactor models. The control objective is that the temperature T evolves within a prespecified performance envelope or ``funnel'' around the set point temperature T*. A simple error feedback control with input constraints of the form u(t)=sat(-k(t)[T(t)-T*] + u*), u* an offset, is introduced which achieves the objective in the presence of disturbances corrupting the measurement. The gain k(t) is a function of the error e(t)=T(t)-T* and its distance to the funnel boundary. The input constraints have to satisfy certain feasibility assumptions in terms of the model data and the operating point T*. |
