Abstract
The paper offers a self-contained account of the theory of first and second order necessary conditions for optimal control problems (with state constraints) based on new principles coming from variational analysis. The key element of the theory is reduction of the problem to unconstrained minimization of a Bolza-type functional with necessarily non-differentiable integrand and off-integral term. This allows to substantially shorten and simplify the proofs and to get new results not detected earlier by traditional variational techniques. This includes a totally new and easily verifiable second order necessary condition for a strong minimum in the classical problem of calculus of variations. The condition is a consequence of a new and more general second order necessary condition for optimal control problems with state constraints. Simple examples show that the new conditions may work when all known necessary conditions fail.
Similar content being viewed by others
Notes
Nonconvex subdifferential calculus, and in particular the “extremal principle” of Kruger–Mordukhovich that can be viewed as a nonconvex extension of the separation theorem, needs closed sets and lsc functions. But the set of trajectories of a control system is typically not closed in the topology of uniform convergence and its closure contains all relaxed trajectories. Therefore the proofs based on nonconvex separation give maximum principle only for relaxed systems—see e.g. [18].
The only work known for me where controllability approach is used to get second order conditions is [1]. But it is assumed in the paper that the optimal control takes value in the interior of the set of admissible controls for all t.
Actually we need a bit less. It will be clear from the proofs that all we need is the existence of a Castaing representation of \(U(\cdot )\) which is a countable collection of measurable selections \(u_i(\cdot ),\ i=1,2,\ldots \) such that for almost every t the closure of the set \(\{u_1(t),u_2(t),\ldots \}\) contains U(t)
In [21] a problem with variable time interval is considered plus the inequality constraints have more general structure than here.
References
Avakov, E.P., Magaril-Il’yaev, G.G.: Controllability and necessary optimality conditions of second order in optimal control. Matem. Sbornik 208, 3–37 (2017) (in Russian; English translation, Sb. Matem. 208 (2017), 585–619)
Buttazzo, G.: Semicontinuity, relaxation and integral representation in the calculus of variations. Pitman Resarch Notes in Mathematics, vol. 207. Pitman (1989)
Castaing, C., Valadier, M.: Convex analysis and measurable multifunctions. Lecture Notes in Mathematics, vol. 580. Springer (1977)
Clarke, F.H.: The maximum principle under minimal hypotheses. SIAM J. Contol Optim. 14, 1078–1091 (1976)
Clarke, F.H.: Optimization and Nonsmooth Analysis. Wiley, Hoboken (1983)
Dubovitzkii, A.Y., Milyutin, A.A.: Problems for extremum under constraints. Zh. Vychisl. Matematiki i Mat. Fiziki 5(3), 395–453 (1965) (in Russian; English translation, USSR Comput. Math. Math. Physics, 5 (1965))
Frankowska, H., Hoehener, D.: Pointwise second order necessary optimality conditions and second order sensitivity relations in optimal control. J. Differ. Equ. 262, 5735–5772 (2017)
Frankowska, H., Osmolovskii, N.P.: Strong local minimizers in optimal control. Problems with state constraints: second order necessary conditions. SIAM J. Control Optim. 56, 2353–2376 (2018)
Gamkrelidze, R.V.: On some extremal problems in the theory of differential equations with applications to the theory of optimal control. SIAM J. Control 3, 106–128 (1965)
Ioffe, A.D.: Necessary and sufficient conditions for a local minimum 1–3. SIAM J. Control Optim. 17, 245–288 (1979)
Ioffe, A.D.: Necessary conditions in nonsmooth optimization. Math. Oper. Res. 9, 159–189 (1984)
Ioffe, A.D.: Euler-Lagrange and Hamiltonian formalisms in dynamic optimization. Trans. Am. Math. Soc. 349, 2871–2900 (1997)
Ioffe, A.D.: Variational Analysis of Regular Mappings. Springer, Berlin (2017)
Ioffe, A.D.: On generalized Bolza problem and its application to dynamic optimization. J. Optim. Theory Appl. 182, 285–309 (2019)
Ioffe, A.D.: Elementary proof of the Pontryagin maximum principle. Vietnam J. Math. https://doi.org/10.1007/s10013-020-00397-0
Loewen, P.D.: Optimal control via nonsmooth analysis. CRM Proceedings and Lecture Notes, vol. 2. AMS (1993)
Milyutin, A.A., Osmolovskii, N.P.: Calculus of Variations and Optimal Control. AMS, Providence (1998)
Mordukhovich, B.S.: Variational Analysis and Generalized Differentiation, vol. 2. Springer, Berlin (2006)
Osmolovskii, N.P.: Necessary quadratic conditions of extremum for discontinuous control in optimal control problems with mixed constraints. J. Math. Sci. 183, 435–577 (2012)
Osmolovskii, N.P.: Necessary second-order conditions for a strong local minimum in a problem with endpoint and control constraints. J. Optim. Theory Appl. 185, 1–16 (2020)
Pales, Z., Zeidan, V.: First and second order optimality conditions in optimal control with pure state constraints. Nonlinear Anal. TMA 67, 2506–2526 (2007)
Pontryagin, L.S., Boltyanskii, V.G., Gamkrelidze, R.V., Mishchenko, E.F.: The Mathemetical Theory of Optimal Processes, Fizmatgiz 1961. Pergamon Press, Oxford (1964). (in Russian)
Vinter, R.B.: Optimal Control. Birkhauser, Basel (2000)
Vinter, R.B.: The Hamiltonian inclusion for nonconvex velocity sets. SIAM J. Control Optim. 52, 1237–1250 (2014)
Author information
Authors and Affiliations
Corresponding author
Additional information
Communicated by J. Jost.
To the memory of Professor Louis Nirenberg.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Ioffe, A.D. Towards the theory of strong minimum in calculus of variations and optimal control: a view from variational analysis. Calc. Var. 59, 83 (2020). https://doi.org/10.1007/s00526-020-01736-2
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00526-020-01736-2