Risk-sensitive mean-field games

Tembine Hamidou, Quanyan Zhu, Tamer Başar

Research output: Contribution to journalArticle

Abstract

In this paper, we study a class of risk-sensitive mean-field stochastic differential games. We show that under appropriate regularity conditions, the mean-field value of the stochastic differential game with exponentiated integral cost functional coincides with the value function satisfying a Hamilton -Jacobi- Bellman (HJB) equation with an additional quadratic term. We provide an explicit solution of the mean-field best response when the instantaneous cost functions are log-quadratic and the state dynamics are affine in the control. An equivalent mean-field risk-neutral problem is formulated and the corresponding mean-field equilibria are characterized in terms of backward-forward macroscopic McKean-Vlasov equations, Fokker-Planck-Kolmogorov equations, and HJB equations. We provide numerical examples on the mean field behavior to illustrate both linear and McKean-Vlasov dynamics.

Original languageEnglish (US)
Article number6656891
Pages (from-to)835-850
Number of pages16
JournalIEEE Transactions on Automatic Control
Volume59
Issue number4
DOIs
StatePublished - 2014

Fingerprint

Vlasov equation
Fokker Planck equation
Cost functions

Keywords

  • Decentralized control
  • H infinity control

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Control and Systems Engineering
  • Computer Science Applications

Cite this

Risk-sensitive mean-field games. / Hamidou, Tembine; Zhu, Quanyan; Başar, Tamer.

In: IEEE Transactions on Automatic Control, Vol. 59, No. 4, 6656891, 2014, p. 835-850.

Research output: Contribution to journalArticle

Hamidou, Tembine ; Zhu, Quanyan ; Başar, Tamer. / Risk-sensitive mean-field games. In: IEEE Transactions on Automatic Control. 2014 ; Vol. 59, No. 4. pp. 835-850.
@article{8480ec4ca60d47369b0df66ca0978ab6,
title = "Risk-sensitive mean-field games",
abstract = "In this paper, we study a class of risk-sensitive mean-field stochastic differential games. We show that under appropriate regularity conditions, the mean-field value of the stochastic differential game with exponentiated integral cost functional coincides with the value function satisfying a Hamilton -Jacobi- Bellman (HJB) equation with an additional quadratic term. We provide an explicit solution of the mean-field best response when the instantaneous cost functions are log-quadratic and the state dynamics are affine in the control. An equivalent mean-field risk-neutral problem is formulated and the corresponding mean-field equilibria are characterized in terms of backward-forward macroscopic McKean-Vlasov equations, Fokker-Planck-Kolmogorov equations, and HJB equations. We provide numerical examples on the mean field behavior to illustrate both linear and McKean-Vlasov dynamics.",
keywords = "Decentralized control, H infinity control",
author = "Tembine Hamidou and Quanyan Zhu and Tamer Başar",
year = "2014",
doi = "10.1109/TAC.2013.2289711",
language = "English (US)",
volume = "59",
pages = "835--850",
journal = "IEEE Transactions on Automatic Control",
issn = "0018-9286",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

TY - JOUR

T1 - Risk-sensitive mean-field games

AU - Hamidou, Tembine

AU - Zhu, Quanyan

AU - Başar, Tamer

PY - 2014

Y1 - 2014

N2 - In this paper, we study a class of risk-sensitive mean-field stochastic differential games. We show that under appropriate regularity conditions, the mean-field value of the stochastic differential game with exponentiated integral cost functional coincides with the value function satisfying a Hamilton -Jacobi- Bellman (HJB) equation with an additional quadratic term. We provide an explicit solution of the mean-field best response when the instantaneous cost functions are log-quadratic and the state dynamics are affine in the control. An equivalent mean-field risk-neutral problem is formulated and the corresponding mean-field equilibria are characterized in terms of backward-forward macroscopic McKean-Vlasov equations, Fokker-Planck-Kolmogorov equations, and HJB equations. We provide numerical examples on the mean field behavior to illustrate both linear and McKean-Vlasov dynamics.

AB - In this paper, we study a class of risk-sensitive mean-field stochastic differential games. We show that under appropriate regularity conditions, the mean-field value of the stochastic differential game with exponentiated integral cost functional coincides with the value function satisfying a Hamilton -Jacobi- Bellman (HJB) equation with an additional quadratic term. We provide an explicit solution of the mean-field best response when the instantaneous cost functions are log-quadratic and the state dynamics are affine in the control. An equivalent mean-field risk-neutral problem is formulated and the corresponding mean-field equilibria are characterized in terms of backward-forward macroscopic McKean-Vlasov equations, Fokker-Planck-Kolmogorov equations, and HJB equations. We provide numerical examples on the mean field behavior to illustrate both linear and McKean-Vlasov dynamics.

KW - Decentralized control

KW - H infinity control

UR - http://www.scopus.com/inward/record.url?scp=84897443281&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84897443281&partnerID=8YFLogxK

U2 - 10.1109/TAC.2013.2289711

DO - 10.1109/TAC.2013.2289711

M3 - Article

VL - 59

SP - 835

EP - 850

JO - IEEE Transactions on Automatic Control

JF - IEEE Transactions on Automatic Control

SN - 0018-9286

IS - 4

M1 - 6656891

ER -