ROM-based multiobjective optimization of elliptic PDEs via numerical continuation

Bennet Gebken; Michael Dellnitz; Sebastian Peitz; Stefan Banholzer; Stefan Volkwein

arxiv: 1906.09075 · v1 · pith:OHZQCR36new · submitted 2019-06-21 · 🧮 math.OC

ROM-based multiobjective optimization of elliptic PDEs via numerical continuation

Stefan Banholzer , Bennet Gebken , Michael Dellnitz , Sebastian Peitz , Stefan Volkwein This is my paper

Pith reviewed 2026-05-25 18:47 UTC · model grok-4.3

classification 🧮 math.OC

keywords multiobjective optimizationreduced basis methodnumerical continuationelliptic PDEPareto setmodel reductionoptimal control

0 comments

The pith

Combining reduced basis surrogates with numerical continuation computes Pareto sets for elliptic PDE multiobjective problems with any number of objectives.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a method for computing the set of optimal compromises in problems where multiple objectives must be balanced and the objectives depend on solutions to elliptic partial differential equations. It replaces repeated full-order PDE solves with a reduced basis surrogate and traces the Pareto set via a continuation procedure that relies on inexact gradient information. This removes the restriction of earlier surrogate-based techniques to only two objectives and lowers overall runtime. A sympathetic reader cares because many design and control tasks require trading off several conflicting criteria inside PDE models, and the high cost of each PDE evaluation has previously made such problems intractable.

Core claim

The central claim is that the reduced basis model reduction method, when combined with a continuation approach that uses inexact gradients, can handle an arbitrary number of objectives in multiobjective optimization problems governed by elliptic PDEs while producing a significant reduction in computing time relative to full-order evaluations.

What carries the argument

The reduced basis surrogate model embedded inside a numerical continuation loop that employs inexact gradients to trace the Pareto set.

If this is right

The method scales to optimization problems that involve more than two objectives.
Computing time drops because most evaluations use the cheap surrogate rather than the full PDE.
Inexact gradient information is sufficient to drive the continuation along the Pareto front.
The approach applies directly to elliptic PDE-constrained optimal control problems.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

If the surrogate accuracy persists, the same continuation strategy could be tested on parabolic or nonlinear PDEs.
Error estimates already available for reduced basis approximations might be turned into rigorous bounds on the distance between the surrogate Pareto set and the true one.
The framework could support online or real-time multiobjective decisions once the offline basis construction is complete.

Load-bearing premise

The reduced basis surrogate must remain accurate enough during the entire continuation process that the Pareto set it produces approximates the set obtained from the full-order PDE model.

What would settle it

For a concrete elliptic PDE test problem with three or more objectives, compute the Pareto set once with the reduced-basis continuation method and once with repeated full-order PDE solves; a substantial mismatch between the two sets would falsify the claim.

read the original abstract

Multiobjective optimization plays an increasingly important role in modern applications, where several objectives are often of equal importance. The task in multiobjective optimization and multiobjective optimal control is therefore to compute the set of optimal compromises (the Pareto set) between the conflicting objectives. Since the Pareto set generally consists of an infinite number of solutions, the computational effort can quickly become challenging which is particularly problematic when the objectives are costly to evaluate as is the case for models governed by partial differential equations (PDEs). To decrease the numerical effort to an affordable amount, surrogate models can be used to replace the expensive PDE evaluations. Existing multiobjective optimization methods using model reduction are limited either to low parameter dimensions or to few (ideally two) objectives. In this article, we present a combination of the reduced basis model reduction method with a continuation approach using inexact gradients. The resulting approach can handle an arbitrary number of objectives while yielding a significant reduction in computing time.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Pairs reduced-basis surrogates with inexact-gradient continuation to scale multiobjective PDE optimization past two objectives, but provides no error propagation through the continuation steps.

read the letter

This paper's core move is to take reduced basis approximation for the elliptic PDE and feed it into a numerical continuation scheme that uses inexact gradients, so the method can trace Pareto sets for any number of objectives instead of stopping at two or low-dimensional cases. That combination directly targets the scaling bottleneck mentioned in the abstract and prior literature limits. It is a straightforward methodological extension rather than a new theoretical device, and the abstract is clear about what existing approaches cannot do. Credit is due for identifying the practical constraint and for keeping the surrogate cheap enough to make continuation feasible. The soft spot is exactly the one the stress-test flags: there is no a posteriori bound that carries the reduced-basis error through the predictor-corrector steps or the multi-objective KKT system, and nothing shows that the computed front converges to the full-order front as the basis tolerance goes to zero. For more than two objectives the manifold dimension increases, so any fixed approximation error can accumulate without being caught. The abstract claims a significant time reduction and that the approach works, but without numerical checks or convergence statements visible here it is impossible to tell whether the surrogate stays accurate enough for the Pareto set itself. This is aimed at researchers already using reduced basis or continuation methods on PDE-constrained problems who need to move beyond two objectives. A reader who knows both tools will see the engineering value quickly. It deserves a serious referee because the scaling claim is practically relevant and the combination is new enough to warrant checking the missing error analysis and the experiments. I would send it out, but with a request for the propagation results and validation that the approximated front stays close to the true one.

Referee Report

2 major / 1 minor

Summary. The paper proposes combining reduced-basis (ROM) surrogate models for elliptic PDEs with a numerical continuation method that employs inexact gradients to approximate the Pareto set in multiobjective optimization problems. The central claim is that this combination handles an arbitrary number of objectives while achieving a substantial reduction in computing time relative to full-order models, extending beyond prior ROM-based methods limited to low parameter dimensions or two objectives.

Significance. If the ROM accuracy is controlled throughout the continuation, the approach would enable tractable computation of high-dimensional Pareto fronts for PDE-constrained problems, a setting where direct methods become prohibitive. The work assembles established techniques (ROM, continuation, inexact gradients) into a practical algorithm; its value would lie in the demonstrated computational savings and extensibility to n>2 objectives, provided the approximation quality is rigorously justified.

major comments (2)

[§3] §3: The continuation procedure (predictor-corrector steps on the reduced KKT system) is formulated entirely on the ROM without an a posteriori error bound that accounts for the effect of the reduced-basis approximation on the traced manifold or on the inexact gradients.
[§4] §4: The derivation of the inexact gradient for the multi-objective problem does not propagate the RB residual or coercivity constants through the continuation steps; consequently there is no theorem establishing that the computed Pareto set converges to the full-order set as the RB tolerance tends to zero, especially when the manifold dimension grows with the number of objectives.

minor comments (1)

The abstract states that the method yields a 'significant reduction in computing time' but supplies no quantitative timing or error data; such results should appear in the main text or a dedicated numerical section.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive and detailed comments. We address each major comment point by point below, indicating planned revisions where appropriate.

read point-by-point responses

Referee: §3: The continuation procedure (predictor-corrector steps on the reduced KKT system) is formulated entirely on the ROM without an a posteriori error bound that accounts for the effect of the reduced-basis approximation on the traced manifold or on the inexact gradients.

Authors: The continuation is performed on the reduced KKT system, with accuracy controlled via standard reduced-basis a posteriori estimators on the state and adjoint solutions that underpin the inexact gradients. A dedicated bound on the traced manifold itself is not derived. In the revised version we will insert a clarifying paragraph in Section 3 on the role of these estimators during continuation and add numerical tests that vary the RB tolerance to quantify its influence on the computed Pareto set. revision: partial
Referee: §4: The derivation of the inexact gradient for the multi-objective problem does not propagate the RB residual or coercivity constants through the continuation steps; consequently there is no theorem establishing that the computed Pareto set converges to the full-order set as the RB tolerance tends to zero, especially when the manifold dimension grows with the number of objectives.

Authors: The inexact gradient is obtained directly from the reduced optimality system; the standard RB residual and coercivity estimators are available but are not propagated analytically through the predictor-corrector steps. The manuscript does not contain a convergence theorem for the Pareto set as the tolerance tends to zero. Its focus is the algorithmic combination and demonstrated computational savings for arbitrary numbers of objectives rather than a full theoretical convergence analysis. revision: no

standing simulated objections not resolved

A theorem establishing that the computed Pareto set converges to the full-order set as the RB tolerance tends to zero, especially when the manifold dimension grows with the number of objectives.

Circularity Check

0 steps flagged

No circularity; methodological combination of established ROM and continuation techniques

full rationale

The paper presents a combination of reduced-basis surrogate modeling with a numerical continuation method using inexact gradients to trace Pareto sets for multiobjective PDE-constrained optimization. The abstract and available text describe this as an application of existing techniques (ROM for cost reduction, continuation for manifold tracing) rather than a closed derivation. No equation is shown to reduce to a self-definition, a fitted parameter renamed as a prediction, or a load-bearing self-citation chain. The central claim of handling arbitrary objectives with reduced compute time rests on the empirical performance of the combined algorithm, which remains externally falsifiable against full-order models and does not rely on internal redefinition of its inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard domain assumptions about elliptic PDEs and reduced basis accuracy; no free parameters or invented entities are mentioned in the abstract.

axioms (1)

domain assumption Elliptic PDEs admit solutions that can be approximated by reduced basis methods with controllable error.
Invoked implicitly when claiming the surrogate replaces expensive PDE evaluations.

pith-pipeline@v0.9.0 · 5699 in / 1039 out tokens · 21560 ms · 2026-05-25T18:47:14.183222+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

30 extracted references · 30 canonical work pages

[1]

Banholzer, D

S. Banholzer, D. Beermann, and S. Volkwein. POD-Based Bicriterial Optimal Control by the Reference Point Method. IFAC-PapersOnLine, 49(8):210–215, 2016

work page 2016
[2]

Banholzer, D

S. Banholzer, D. Beermann, and S. Volkwein. POD-Based Error Control for Reduced-Order Bicriterial PDE-Constrained Optimization. Annual Reviews in Control , 44:226–237, 2017

work page 2017
[3]

Beermann, M

D. Beermann, M. Dellnitz, S. Peitz, and S. Volkwein. POD-based multiobjective optimal control of PDEs with non-smooth objectives. In Proceedings in Applied Mathematics and Mechanics (PAMM), pages 51–54, 2017

work page 2017
[4]

Beermann, M

D. Beermann, M. Dellnitz, S. Peitz, and S. Volkwein. Set-Oriented Multiobjective Optimal Control of PDEs using Proper Orthogonal Decomposition. In Reduced-Order Modeling (ROM) for Simulation and Optimization , pages 47–72. Springer, 2018

work page 2018
[5]

A. Buﬀa, Y. Maday, A. T. Patera, C. Prudhomme, and G. Turinici. A priori convergence of the greedy algorithm for the parametrized reduced basis method. ESAIM: Mathematical Modelling and Numerical Analysis , 46(3):595–603, 2012

work page 2012
[6]

Chugh, K

T. Chugh, K. Sindhya, J. Hakanen, and K. Miettinen. A survey on handling computa- tionally expensive multiobjective optimization problems with evolutionary algorithms. Soft Computing, pages 1–30, 2017

work page 2017
[7]

C. A. Coello Coello, G. B. Lamont, and D. A. Van Veldhuizen. Evolutionary Algorithms for Solving Multi-Objective Problems , volume 2. Springer Science & Business Media, 2007

work page 2007
[8]

Dellnitz, O

M. Dellnitz, O. Sch¨ utze, and T. Hestermeyer. Covering Pareto Sets by Multilevel Subdi- vision Techniques. Journal of Optimization Theory and Applications , 124(1):113–136, Jan 2005

work page 2005
[9]

M. Ehrgott. Multicriteria optimization . Springer Berlin Heidelberg New York, 2 edition, 2005

work page 2005
[10]

Gebken, S

B. Gebken, S. Peitz, and M. Dellnitz. On the hierarchical structure of Pareto critical sets. Journal of Global Optimization , 73(4):891–913, 2019

work page 2019
[11]

M. A. Grepl and A. T. Patera. A posteriori error bounds for reduced-basis approximations of parametrized parabolic partial diﬀerential equations. ESAIM: Mathematical Modelling and Numerical Analysis, 39(1):157–181, 2005

work page 2005
[12]

Haasdonk, J

B. Haasdonk, J. Salomon, and B. Wohlmuth. A reduced basis method for the simulation of american options. In Numerical Mathematics and Advanced Applications 2011 , pages 821–829, Berlin, Heidelberg, 2013. Springer Berlin Heidelberg

work page 2011
[13]

Hesthaven, G

J. Hesthaven, G. Rozza, and B. Stamm. Certiﬁed reduced basis methods for parametrized partial diﬀerential equations. SpringerBriefs in Mathematics , 2016

work page 2016
[14]

Hillermeier

C. Hillermeier. Nonlinear Multiobjective Optimization: A Generalized Homotopy Approach . Birkh¨ auser Basel, 2001

work page 2001
[15]

Hinze, R

M. Hinze, R. Pinnau, M. Ulbrich, and S. Ulbrich. Optimization with PDE Constraints . Springer, 2009. 21

work page 2009
[16]

Iapichino, S

L. Iapichino, S. Trenz, and S. Volkwein. Multiobjective optimal control of semilinear parabolic problems using POD. In B. Karas¨ ozen, M. Manguoglu, M. Tezer-Sezgin, S. Gok- tepe, and ¨O. Ugur, editors, Numerical Mathematics and Advanced Applications (ENU- MATH 2015), pages 389–397. Springer, 2016

work page 2015
[17]

Iapichino, S

L. Iapichino, S. Ulbrich, and S. Volkwein. Multiobjective PDE-Constrained Optimization Using the Reduced-Basis Method. Advances in Computational Mathematics, 43(5):945–972, 2017

work page 2017
[18]

Kunisch and S

K. Kunisch and S. Volkwein. Galerkin proper orthogonal decomposition methods for parabolic problems. Numerische Mathematik, 90(1):117–148, 2001

work page 2001
[19]

J. Lee. Introduction to Smooth Manifolds . Springer-Verlag New York, 2012

work page 2012
[20]

Ohlberger and F

M. Ohlberger and F. Schindler. Non-conforming localized model reduction with online enrichment: Towards optimal complexity in pde constrained optimization. In Finite Vol- umes for Complex Applications VIII - Hyperbolic, Elliptic and Parabolic Problems , pages 357–365. Springer International Publishing, 2017

work page 2017
[21]

Peitz and M

S. Peitz and M. Dellnitz. Gradient-Based Multiobjective Optimization with Uncertainties , pages 159–182. Springer International Publishing, 2017

work page 2017
[22]

Peitz and M

S. Peitz and M. Dellnitz. A Survey of Recent Trends in Multiobjective Optimal Control Surrogate Models, Feedback Control and Objective Reduction. Mathematical and Compu- tational Applications, 23(2), 2018

work page 2018
[23]

Peitz, S

S. Peitz, S. Ober-Bl¨ obaum, and M. Dellnitz. Multiobjective Optimal Control Methods for the Navier-Stokes Equations Using Reduced Order Modeling. Acta Applicandae Mathemat- icae, 161(1):171–199, 2019

work page 2019
[24]

Peitz, K

S. Peitz, K. Sch¨ afer, S. Ober-Bl¨ obaum, J. Eckstein, U. K¨ ohler, and M. Dellnitz. A Multiob- jective MPC Approach for Autonomously Driven Electric Vehicles. IFAC PapersOnLine, 50(1):8674–8679, 2017

work page 2017
[25]

Quarteroni, A

A. Quarteroni, A. Manoni, and F. Negri. Reduced Basis Methods for Partial Diﬀerential Equations. Springer, 2016

work page 2016
[26]

Rozza, D

G. Rozza, D. B. P. Huynh, and A. T. Patera. Reduced basis approximation and a posteriori error estimation for aﬃnely parametrized elliptic coercive partial diﬀerential equations. Archives of Computational Methods in Engineering , 15(3):1, 2007

work page 2007
[27]

Sch¨ utze, O

O. Sch¨ utze, O. Cuate, A. Mart´ ın, S. Peitz, and M. Dellnitz. Pareto explorer: a global/local exploration tool for many-objective optimization problems. Engineering Optimization , pages 1–24, 05 2019

work page 2019
[28]

Sch¨ utze, A

O. Sch¨ utze, A. Dell’Aere, and M. Dellnitz. On Continuation Methods for the Numerical Treatment of Multi-Objective Optimization Problems. In Practical Approaches to Multi- Objective Optimization, number 04461 in Dagstuhl Seminar Proceedings, Dagstuhl, Ger- many, 2005. Internationales Begegnungs- und Forschungszentrum f¨ ur Informatik (IBFI), Schloss Dags...

work page 2005
[29]

Sirovich

L. Sirovich. Turbulence and the dynamics of coherent structures part I: coherent structures. Quarterly of Applied Mathematics , XLV(3):561–571, 1987

work page 1987
[30]

Non-smooth and Complementarity- based Distributed Parameter Systems

M. Tabatabaei, J. Hakanen, M. Hartikainen, K. Miettinen, and K. Sindhya. A survey on handling computationally expensive multiobjective optimization problems using surrogates: non-nature inspired methods. Structural and Multidisciplinary Optimization , 52(1):1–25, 2015. Acknowledgment This research was funded by the DFG Priority Programme 1962 “Non-smooth ...

work page 2015

[1] [1]

Banholzer, D

S. Banholzer, D. Beermann, and S. Volkwein. POD-Based Bicriterial Optimal Control by the Reference Point Method. IFAC-PapersOnLine, 49(8):210–215, 2016

work page 2016

[2] [2]

Banholzer, D

S. Banholzer, D. Beermann, and S. Volkwein. POD-Based Error Control for Reduced-Order Bicriterial PDE-Constrained Optimization. Annual Reviews in Control , 44:226–237, 2017

work page 2017

[3] [3]

Beermann, M

D. Beermann, M. Dellnitz, S. Peitz, and S. Volkwein. POD-based multiobjective optimal control of PDEs with non-smooth objectives. In Proceedings in Applied Mathematics and Mechanics (PAMM), pages 51–54, 2017

work page 2017

[4] [4]

Beermann, M

D. Beermann, M. Dellnitz, S. Peitz, and S. Volkwein. Set-Oriented Multiobjective Optimal Control of PDEs using Proper Orthogonal Decomposition. In Reduced-Order Modeling (ROM) for Simulation and Optimization , pages 47–72. Springer, 2018

work page 2018

[5] [5]

A. Buﬀa, Y. Maday, A. T. Patera, C. Prudhomme, and G. Turinici. A priori convergence of the greedy algorithm for the parametrized reduced basis method. ESAIM: Mathematical Modelling and Numerical Analysis , 46(3):595–603, 2012

work page 2012

[6] [6]

Chugh, K

T. Chugh, K. Sindhya, J. Hakanen, and K. Miettinen. A survey on handling computa- tionally expensive multiobjective optimization problems with evolutionary algorithms. Soft Computing, pages 1–30, 2017

work page 2017

[7] [7]

C. A. Coello Coello, G. B. Lamont, and D. A. Van Veldhuizen. Evolutionary Algorithms for Solving Multi-Objective Problems , volume 2. Springer Science & Business Media, 2007

work page 2007

[8] [8]

Dellnitz, O

M. Dellnitz, O. Sch¨ utze, and T. Hestermeyer. Covering Pareto Sets by Multilevel Subdi- vision Techniques. Journal of Optimization Theory and Applications , 124(1):113–136, Jan 2005

work page 2005

[9] [9]

M. Ehrgott. Multicriteria optimization . Springer Berlin Heidelberg New York, 2 edition, 2005

work page 2005

[10] [10]

Gebken, S

B. Gebken, S. Peitz, and M. Dellnitz. On the hierarchical structure of Pareto critical sets. Journal of Global Optimization , 73(4):891–913, 2019

work page 2019

[11] [11]

M. A. Grepl and A. T. Patera. A posteriori error bounds for reduced-basis approximations of parametrized parabolic partial diﬀerential equations. ESAIM: Mathematical Modelling and Numerical Analysis, 39(1):157–181, 2005

work page 2005

[12] [12]

Haasdonk, J

B. Haasdonk, J. Salomon, and B. Wohlmuth. A reduced basis method for the simulation of american options. In Numerical Mathematics and Advanced Applications 2011 , pages 821–829, Berlin, Heidelberg, 2013. Springer Berlin Heidelberg

work page 2011

[13] [13]

Hesthaven, G

J. Hesthaven, G. Rozza, and B. Stamm. Certiﬁed reduced basis methods for parametrized partial diﬀerential equations. SpringerBriefs in Mathematics , 2016

work page 2016

[14] [14]

Hillermeier

C. Hillermeier. Nonlinear Multiobjective Optimization: A Generalized Homotopy Approach . Birkh¨ auser Basel, 2001

work page 2001

[15] [15]

Hinze, R

M. Hinze, R. Pinnau, M. Ulbrich, and S. Ulbrich. Optimization with PDE Constraints . Springer, 2009. 21

work page 2009

[16] [16]

Iapichino, S

L. Iapichino, S. Trenz, and S. Volkwein. Multiobjective optimal control of semilinear parabolic problems using POD. In B. Karas¨ ozen, M. Manguoglu, M. Tezer-Sezgin, S. Gok- tepe, and ¨O. Ugur, editors, Numerical Mathematics and Advanced Applications (ENU- MATH 2015), pages 389–397. Springer, 2016

work page 2015

[17] [17]

Iapichino, S

L. Iapichino, S. Ulbrich, and S. Volkwein. Multiobjective PDE-Constrained Optimization Using the Reduced-Basis Method. Advances in Computational Mathematics, 43(5):945–972, 2017

work page 2017

[18] [18]

Kunisch and S

K. Kunisch and S. Volkwein. Galerkin proper orthogonal decomposition methods for parabolic problems. Numerische Mathematik, 90(1):117–148, 2001

work page 2001

[19] [19]

J. Lee. Introduction to Smooth Manifolds . Springer-Verlag New York, 2012

work page 2012

[20] [20]

Ohlberger and F

M. Ohlberger and F. Schindler. Non-conforming localized model reduction with online enrichment: Towards optimal complexity in pde constrained optimization. In Finite Vol- umes for Complex Applications VIII - Hyperbolic, Elliptic and Parabolic Problems , pages 357–365. Springer International Publishing, 2017

work page 2017

[21] [21]

Peitz and M

S. Peitz and M. Dellnitz. Gradient-Based Multiobjective Optimization with Uncertainties , pages 159–182. Springer International Publishing, 2017

work page 2017

[22] [22]

Peitz and M

S. Peitz and M. Dellnitz. A Survey of Recent Trends in Multiobjective Optimal Control Surrogate Models, Feedback Control and Objective Reduction. Mathematical and Compu- tational Applications, 23(2), 2018

work page 2018

[23] [23]

Peitz, S

S. Peitz, S. Ober-Bl¨ obaum, and M. Dellnitz. Multiobjective Optimal Control Methods for the Navier-Stokes Equations Using Reduced Order Modeling. Acta Applicandae Mathemat- icae, 161(1):171–199, 2019

work page 2019

[24] [24]

Peitz, K

S. Peitz, K. Sch¨ afer, S. Ober-Bl¨ obaum, J. Eckstein, U. K¨ ohler, and M. Dellnitz. A Multiob- jective MPC Approach for Autonomously Driven Electric Vehicles. IFAC PapersOnLine, 50(1):8674–8679, 2017

work page 2017

[25] [25]

Quarteroni, A

A. Quarteroni, A. Manoni, and F. Negri. Reduced Basis Methods for Partial Diﬀerential Equations. Springer, 2016

work page 2016

[26] [26]

Rozza, D

G. Rozza, D. B. P. Huynh, and A. T. Patera. Reduced basis approximation and a posteriori error estimation for aﬃnely parametrized elliptic coercive partial diﬀerential equations. Archives of Computational Methods in Engineering , 15(3):1, 2007

work page 2007

[27] [27]

Sch¨ utze, O

O. Sch¨ utze, O. Cuate, A. Mart´ ın, S. Peitz, and M. Dellnitz. Pareto explorer: a global/local exploration tool for many-objective optimization problems. Engineering Optimization , pages 1–24, 05 2019

work page 2019

[28] [28]

Sch¨ utze, A

O. Sch¨ utze, A. Dell’Aere, and M. Dellnitz. On Continuation Methods for the Numerical Treatment of Multi-Objective Optimization Problems. In Practical Approaches to Multi- Objective Optimization, number 04461 in Dagstuhl Seminar Proceedings, Dagstuhl, Ger- many, 2005. Internationales Begegnungs- und Forschungszentrum f¨ ur Informatik (IBFI), Schloss Dags...

work page 2005

[29] [29]

Sirovich

L. Sirovich. Turbulence and the dynamics of coherent structures part I: coherent structures. Quarterly of Applied Mathematics , XLV(3):561–571, 1987

work page 1987

[30] [30]

Non-smooth and Complementarity- based Distributed Parameter Systems

M. Tabatabaei, J. Hakanen, M. Hartikainen, K. Miettinen, and K. Sindhya. A survey on handling computationally expensive multiobjective optimization problems using surrogates: non-nature inspired methods. Structural and Multidisciplinary Optimization , 52(1):1–25, 2015. Acknowledgment This research was funded by the DFG Priority Programme 1962 “Non-smooth ...

work page 2015