On the Sequential Normal Compactness Condition and its Restrictiveness in Selected Function Spaces


Sequential normal compactness is one of the most important properties in terms of modern variational analysis. It is necessary for the derivation of calculus rules for the computation of generalized normals to set intersections or preimages of sets under transformations. While sequential normal compactness is inherent in finite-dimensional Banach spaces, its presence has to be checked in the infinite-dimensional situation. In this paper, we show that broad classes of sets in Lebesgue and Sobolev spaces which are reasonable in the context of optimal control suffer from an intrinsic lack of sequential normal compactness.

This is a preview of subscription content, log in to check access.


  1. 1.

    Adams, R.A., Fournier, J.J.F.: Sobolev spaces. Elsevier Science, Oxford (2003)

    Google Scholar 

  2. 2.

    Asplund, E.: Fréchet differentiability of convex functions. Acta Math. 121, 31–47 (1968). https://doi.org/10.1007/BF02391908

    MathSciNet  Article  MATH  Google Scholar 

  3. 3.

    Attouch, H., Buttazzo, G., Michaille, G.: Variational analysis in Sobolev and BV spaces, volume 6 of MPS/SIAM Series on Optimization. Society for Industrial and Applied Mathematics (SIAM), Philadelphia (2006)

    Google Scholar 

  4. 4.

    Aubin, J.-P., Frankowska, H.: Set-valued analysis. Modern Birkhäuser Classics. Birkhäuser Boston Inc., Boston (2009). ISBN 978-0-8176-4847-3. Reprint of the 1990 edition

    Google Scholar 

  5. 5.

    Bogachev, V.I.: Measure Theory. Springer, Berlin (2007)

    Google Scholar 

  6. 6.

    Borwein, J.M., Strojwas, H.M.: Tangential approximations. Nonlinear Anal. Theory Methods Appl. 9(12), 1347–1366 (1985). https://doi.org/10.1016/0362-546X(85)90095-1

    MathSciNet  Article  MATH  Google Scholar 

  7. 7.

    Borwein, J.M., Lucet, Y., Mordukhovich, B.S.: Compactly epi-Lipschitzian convex sets and functions in normed spaces. J. Convex Anal. 7(2), 375–393 (2000)

    MathSciNet  MATH  Google Scholar 

  8. 8.

    Fabian, M., Mordukhovich, B.S.: Sequential normal compactness versus topological normal compactness in variational analysis. Nonlinear Anal. Theory Methods Appl. 54 (6), 1057–1067 (2003). https://doi.org/10.1016/S0362-546X(03)00126-3

    MathSciNet  Article  MATH  Google Scholar 

  9. 9.

    Guo, L., Ye, J.J.: Necessary optimality conditions for optimal control problems with equilibrium constraints. SIAM J. Control. Optim. 54(5), 2710–2733 (2016). https://doi.org/10.1137/15M1013493

    MathSciNet  Article  MATH  Google Scholar 

  10. 10.

    Harder, F., Wachsmuth, G.: The limiting normal cone of a complementarity set in Sobolev spaces. Preprint TU Chemnitz, pp. 1–31. https://www.tu-chemnitz.de/mathematik/part_dgl/publications/Harder_Wachsmuth__The_limiting_normal_cone_of_a_complementarity_set_in_Sobolev_spaces.pdf (2017)

  11. 11.

    Hiai, F., Umegaki, H.: Integrals, conditional expectations, and martingales of multivalued functions. J. Multivar. Anal. 7(1), 149–182 (1977). https://doi.org/10.1016/0047-259X(77)90037-9

    MathSciNet  Article  MATH  Google Scholar 

  12. 12.

    Hintermüller, M., Mordukhovich, B.S., Surowiec, T.M.: Several approaches for the derivation of stationarity conditions for elliptic MPECs with upper-level control constraints. Math. Program. 146(1), 555–582 (2014). https://doi.org/10.1007/s10107-013-0704-6

    MathSciNet  Article  MATH  Google Scholar 

  13. 13.

    Ioffe, A.D.: Coderivative compactness, metric regularity and subdifferential calculus. In: Théra, M. (ed.) Experimental, Constructive and Nonlinear Analysis, volume 27 of Canadian Mathematical Society of Conference Proceedings, pp. 123–164. AMS, Providence, RI (1990)

  14. 14.

    Jarušek, J., Outrata, J.V.: On sharp necessary optimality conditions in control of contact problems with strings. Nonlinear Anal. Theory Methods Appl. 67(4), 1117–1128 (2007). https://doi.org/10.1016/j.na.2006.05.021

    MathSciNet  Article  MATH  Google Scholar 

  15. 15.

    Loewen, P.D.: Limits of Fréchet normals in nonsmooth analysis. In: Ioffe, A.D., Marcus, L., Reich, S. (eds.) Optimization and Nonlinear Analysis, volume 244 of Pitman Research Notes Math. Ser., pp. 178–188. Longman, Harlow, Essex, UK (1992)

  16. 16.

    Megginson, R.E.: An introduction to Banach space theory. Graduate texts in mathematics. Springer, New York (1998)

    Google Scholar 

  17. 17.

    Mehlitz, P., Wachsmuth, G.: On the limiting normal cone to pointwise defined sets in Lebesgue spaces. Set-valued and variational analysis (2016). https://doi.org/10.1007/s11228-016-0393-4

  18. 18.

    Mehlitz, P., Wachsmuth, G.: The weak sequential closure of decomposable sets in Lebesgue spaces and its application to variational geometry. Set-valued and variational analysis (2017). https://doi.org/10.1007/s11228-017-0464-1

  19. 19.

    Mordukhovich, B.S.: Generalized differential calculus for nonsmooth and set-valued mappings. J. Math. Anal. Appl. 183(1), 250–288 (1994). https://doi.org/10.1006/jmaa.1994.1144

    MathSciNet  Article  MATH  Google Scholar 

  20. 20.

    Mordukhovich, B.S.: Variational analysis and generalized differentiation. Springer-Verlag, Berlin (2006)

    Google Scholar 

  21. 21.

    Mordukhovich, B.S., Shao, Y.: Nonconvex differential calculus for infinite-dimensional multifunctions. Set-Valued Anal. 4(3), 205–236 (1996a). https://doi.org/10.1007/BF00419366

  22. 22.

    Mordukhovich, B.S., Shao, Y.: Nonsmooth sequential analysis in Asplund spaces. Trans. Amer. Math. Soc. 348, 1235–1280 (1996b). https://doi.org/10.1090/S0002-9947-96-01543-7

  23. 23.

    Mordukhovich, B.S., Shao, Y.: Stability of set-valued mappings in infinite dimensions: point criteria and applications. SIAM J. Control. Optim. 35(1), 285–314 (1997). https://doi.org/10.1137/S0363012994278171

    MathSciNet  Article  MATH  Google Scholar 

  24. 24.

    Mordukhovich, B.S., Wang, B.: Calculus of sequential normal compactness in variational analysis. J. Math. Anal. Appl. 282 (1), 63–84 (2003). https://doi.org/10.1016/S0022-247X(02)00385-2

    MathSciNet  Article  MATH  Google Scholar 

  25. 25.

    Outrata, J., Jarušek, J., Stará, J.: On optimality conditions in control of elliptic variational inequalities. Set-Valued and Variational Anal. 19(1), 23–42 (2011). https://doi.org/10.1007/s11228-010-0158-4

    MathSciNet  Article  MATH  Google Scholar 

  26. 26.

    Outrata, J.V., Römisch, W.: On optimality conditions for some nonsmooth optimization problems over L p spaces. J. Optim. Theory Appl. 126 (2), 411–438 (2005). https://doi.org/10.1007/s10957-005-4724-0

    MathSciNet  Article  MATH  Google Scholar 

  27. 27.

    Papageorgiou, N.S., Kyritsi-Yiallourou, S.T.: Handbook of applied analysis, volume 19 of advances in mechanics and mathematics. Springer, New York. ISBN 978-0-387-78906-4 (2009)

  28. 28.

    Rockafellar, R.T.: Integrals which are convex functionals. Pacific J. Math. 24(3), 525–539 (1968)

    MathSciNet  Article  MATH  Google Scholar 

  29. 29.

    Rockafellar, R.T., Wets, R.J.-B.: Variational analysis, volume 317 of Grundlehren der mathematischen Wissenschaften. Springer, Berlin (1998)

    Google Scholar 

  30. 30.

    Simonnet, M.: Measures and probabilities. Springer, New York (1996)

    Google Scholar 

  31. 31.

    Wachsmuth, G.: Towards M-stationarity for optimal control of the obstacle problem with control constraints. SIAM J. Control. Optim. 54(2), 964–986 (2016). https://doi.org/10.1137/140980582

    MathSciNet  Article  MATH  Google Scholar 

  32. 32.

    Yost, D.: Asplund spaces for beginners. Acta Universitatis Carolinae Mathematica et Physica 34(2), 159–177 (1993)

    MathSciNet  MATH  Google Scholar 

Download references


This work is supported by the DFG grant Analysis and Solution Methods for Bilevel Optimal Control Problems within the Priority Program SPP 1962 (Non-smooth and Complementarity-based Distributed Parameter Systems: Simulation and Hierarchical Optimization).

Author information



Corresponding author

Correspondence to Patrick Mehlitz.



In the lemma below, we show a nearby relationship between the operators lin and conv.

Lemma A.1

Let \(S\subset \mathcal X\) be a nonempty subset of a Banach space \(\mathcal X\) . Then the relation \(\operatorname {lin} S=\operatorname {conv}\bigcup _{\alpha \in \mathbb {R}}\alpha S\) holds true.


Let us prove the inclusion [⊂]first. Since we have \(S\subset \operatorname {conv}\bigcup _{\alpha \in \mathbb {R}}\alpha S\), it is sufficient toshow that \(L:=\operatorname {conv}\bigcup _{\alpha \in \mathbb {R}}\alpha S\) is a linearspace. Therefore, we choose x,yL and γx,γy arbitrarily. Then we find integers n,m,vectors \({s^{x}_{1}},\dots ,{s^{x}_{n}}, {s^{y}_{1}},\dots ,{s^{y}_{m}}\in S\), scalars\({\alpha ^{x}_{1}},\dots ,{\alpha ^{x}_{n}},{\alpha ^{y}_{1}},\dots ,{\alpha ^{y}_{m}}\in \mathbb {R}\), and scalars \(\mu _{1},\dots ,\mu _{n},\nu _{1},\dots ,\nu _{m}\in \mathbb {R} ^{+}_{0}\) such that

$$x=\sum\limits_{i = 1}^{n}\mu_{i}{\alpha^{x}_{i}}{s^{x}_{i}},\quad y=\sum\limits_{j = 1}^{m}\nu_{j}{\alpha^{y}_{j}}{s^{y}_{j}},\quad\sum\limits_{i = 1}^{n}\mu_{i}= 1,\quad\sum\limits_{j = 1}^{m}\nu_{j}= 1. $$

Thus, we obtain

$$\gamma^{x}x+\gamma^{y}y=\sum\limits_{i = 1}^{n}\mu_{i}\gamma^{x}{\alpha^{x}_{i}}{s^{x}_{i}}+\sum\limits_{j = 1}^{m}\nu_{j}\gamma^{y}{\alpha^{y}_{j}}{s^{y}_{j}}=\sum\limits_{i = 1}^{n}\frac{\mu_{i}}{2}2\gamma^{x}{\alpha^{x}_{i}}{s^{x}_{i}}+\sum\limits_{j = 1}^{m}\frac{\nu_{j}}{2}2\gamma^{y}{\alpha^{y}_{j}}{s^{y}_{j}}\in L$$

since\(\tfrac {\mu _{1}}{2},\dots ,\tfrac {\mu _{n}}{2},\tfrac {\nu _{1}}{2},\dots ,\tfrac {\nu _{m}}{2}\in \mathbb {R} ^{+}_{0}\) are nonnegative scalars which satisfy

$$\sum\limits_{i = 1}^{n}\frac{\mu_{i}}{2}+\sum\limits_{j = 1}^{m}\frac{\nu_{j}}{2}= 1.$$

Consequently, L is a linearspace which contains S, i.e. lin SL.

The converse inclusion [⊃]follows from the obvious fact that αS ⊂ lin S holds for all α.This shows L ⊂ conv lin S = lin S andcompletes the proof. □

In this paper, we exploit the following consequence of Lebesgue’s dominated convergence theorem.

Lemma A.2

Fix a complete andσ-finitemeasure space (Ω,Σ, 𝔪)as well as a parameterp ∈ [1,). LetuLp (𝔪; q)be arbitrarily chosen and letk}k ⊂Σbe a sequence of sets which satisfies 𝔪(Ωk) → 0ask. Then\(\chi _{{\Omega }_{k}}u\to 0\)and\((1-\chi _{{\Omega }_{k}})u\to u\)inholdLp (𝔪; q)ask.


By construction, the sequences \(\{\chi _{{\Omega }_{k}}u\}_{k\in \mathbb {N} }\) and\(\{(1-\chi _{{\Omega }_{k}})u\}_{k\in \mathbb {N} }\) converge pointwise almost everywhere on Ωto 0 and u, respectively. Moreover, \(\left \{|{\chi _{{\Omega }_{k}}u}{2}\right \}_{k\in \mathbb {N} }\) and\(\left \{|{(1-\chi _{{\Omega }_{k}})u}{2}\right \}_{k\in \mathbb {N} }\) areboth majorized by |u2 ∈ Lp (𝔪), i.e.

$$\forall k\in\mathbb {N}\,\forall\omega\in{\Omega}\colon\quad|{\chi_{{\Omega}_{k}}(\omega)u(\omega)}{2}\leq|{u(\omega)}{2}\quad|{(1-\chi_{{\Omega}_{k}}(\omega))u(\omega)}{2}\leq|{u(\omega)}{2} $$

is valid. Thus, the lemma’s assertion follows from the dominated convergence theorem, see [30, Theorem5.2.2]. □

The following lemma shows that first-order Sobolev regularity is preserved while taking the component- and pointwise minimum in vector-valued function spaces. It is a straightforward consequence of [3, Corollary 5.8.2].

Lemma A.3

Fixp ∈ (1,). For functionsy,zW1,p(Ω)q, the componentwiseminimum function min{y; z}belongs toW1,p(Ω)qas well.For any indicesi ∈{1,…,q}as well asj ∈{1,…,d}, weobtain

$$\tfrac{\partial}{\partial\omega_{j}}\min\{y;z\}_{i}=\chi_{\{\omega\in{\Omega}\,|\,y_{i}(\omega)<z_{i}(\omega)\}}\tfrac{\partial}{\partial\omega_{j}}y_{i}+\chi_{\{\omega\in{\Omega}\,|\,y_{i}(\omega)\geq z_{i}(\omega)\}}\tfrac{\partial}{\partial \omega_{j}}z_{i}. $$

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Mehlitz, P. On the Sequential Normal Compactness Condition and its Restrictiveness in Selected Function Spaces. Set-Valued Var. Anal 27, 763–782 (2019). https://doi.org/10.1007/s11228-018-0475-6

Download citation


  • Decomposable set
  • Optimal control
  • Sequential normal compactness

Mathematics Subject Classification (2010)

  • 28B05
  • 49J53