On conditional cuts for stochastic dual dynamic programming

Multistage stochastic programs arise in many applications from engineering whenever a set of inventories or stocks has to be valued. Such is the case in seasonal storage valuation of a set of cascaded reservoir chains in hydro management. A popular method is stochastic dual dynamic programming (SDDP...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: W. van Ackooij, X. Warin
Formato: article
Lenguaje:EN
Publicado: Elsevier 2020
Materias:
Acceso en línea:https://doaj.org/article/c48db89d44434951b699b68edbd0446e
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:c48db89d44434951b699b68edbd0446e
record_format dspace
spelling oai:doaj.org-article:c48db89d44434951b699b68edbd0446e2021-12-02T05:01:15ZOn conditional cuts for stochastic dual dynamic programming2192-440610.1007/s13675-020-00123-yhttps://doaj.org/article/c48db89d44434951b699b68edbd0446e2020-06-01T00:00:00Zhttp://www.sciencedirect.com/science/article/pii/S219244062100126Xhttps://doaj.org/toc/2192-4406Multistage stochastic programs arise in many applications from engineering whenever a set of inventories or stocks has to be valued. Such is the case in seasonal storage valuation of a set of cascaded reservoir chains in hydro management. A popular method is stochastic dual dynamic programming (SDDP), especially when the dimensionality of the problem is large and dynamic programming is no longer an option. The usual assumption of SDDP is that uncertainty is stage-wise independent, which is highly restrictive from a practical viewpoint. When possible, the usual remedy is to increase the state-space to account for some degree of dependency. In applications, this may not be possible or it may increase the state-space by too much. In this paper, we present an alternative based on keeping a functional dependency in the SDDP—cuts related to the conditional expectations in the dynamic programming equations. Our method is based on popular methodology in mathematical finance, where it has progressively replaced scenario trees due to superior numerical performance. We demonstrate the interest of combining this way of handling dependency in uncertainty and SDDP on a set of numerical examples. Our method is readily available in the open-source software package StOpt.W. van AckooijX. WarinElsevierarticle90C1565C35Applied mathematics. Quantitative methodsT57-57.97Electronic computers. Computer scienceQA75.5-76.95ENEURO Journal on Computational Optimization, Vol 8, Iss 2, Pp 173-199 (2020)
institution DOAJ
collection DOAJ
language EN
topic 90C15
65C35
Applied mathematics. Quantitative methods
T57-57.97
Electronic computers. Computer science
QA75.5-76.95
spellingShingle 90C15
65C35
Applied mathematics. Quantitative methods
T57-57.97
Electronic computers. Computer science
QA75.5-76.95
W. van Ackooij
X. Warin
On conditional cuts for stochastic dual dynamic programming
description Multistage stochastic programs arise in many applications from engineering whenever a set of inventories or stocks has to be valued. Such is the case in seasonal storage valuation of a set of cascaded reservoir chains in hydro management. A popular method is stochastic dual dynamic programming (SDDP), especially when the dimensionality of the problem is large and dynamic programming is no longer an option. The usual assumption of SDDP is that uncertainty is stage-wise independent, which is highly restrictive from a practical viewpoint. When possible, the usual remedy is to increase the state-space to account for some degree of dependency. In applications, this may not be possible or it may increase the state-space by too much. In this paper, we present an alternative based on keeping a functional dependency in the SDDP—cuts related to the conditional expectations in the dynamic programming equations. Our method is based on popular methodology in mathematical finance, where it has progressively replaced scenario trees due to superior numerical performance. We demonstrate the interest of combining this way of handling dependency in uncertainty and SDDP on a set of numerical examples. Our method is readily available in the open-source software package StOpt.
format article
author W. van Ackooij
X. Warin
author_facet W. van Ackooij
X. Warin
author_sort W. van Ackooij
title On conditional cuts for stochastic dual dynamic programming
title_short On conditional cuts for stochastic dual dynamic programming
title_full On conditional cuts for stochastic dual dynamic programming
title_fullStr On conditional cuts for stochastic dual dynamic programming
title_full_unstemmed On conditional cuts for stochastic dual dynamic programming
title_sort on conditional cuts for stochastic dual dynamic programming
publisher Elsevier
publishDate 2020
url https://doaj.org/article/c48db89d44434951b699b68edbd0446e
work_keys_str_mv AT wvanackooij onconditionalcutsforstochasticdualdynamicprogramming
AT xwarin onconditionalcutsforstochasticdualdynamicprogramming
_version_ 1718400852359643136