Data‐Driven Stochastic Game Theoretic Differential Dynamic Programming

ABSTRACT This paper introduces a novel approach for designing optimal control using data‐driven Stochastic Game Theoretic Differential Dynamic Programming (SGT‐DDP). The proposed method addresses unknown stochastic systems by approximating both drift and diffusion dynamics. The drift dynamics is est...

Full description

Saved in:
Bibliographic Details
Published in:International journal of robust and nonlinear control Vol. 35; no. 13; pp. 5343 - 5354
Main Authors: Sarbaz, Mohammad, Sun, Wei
Format: Journal Article
Language:English
Published: Hoboken, USA John Wiley & Sons, Inc 10.09.2025
Wiley Subscription Services, Inc
Subjects:
Approximation
differential dynamic programming
Drift estimation
Dynamic programming
Dynamical systems
Game theory
Gaussian process
gaussian process regression
Nonlinear systems
Optimal control
stochastic dynamics
Stochastic systems
ISSN:1049-8923, 1099-1239
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:ABSTRACT This paper introduces a novel approach for designing optimal control using data‐driven Stochastic Game Theoretic Differential Dynamic Programming (SGT‐DDP). The proposed method addresses unknown stochastic systems by approximating both drift and diffusion dynamics. The drift dynamics is estimated via Gaussian Process Regression (GPR) using input–output data. The diffusion dynamics is approximated from the noise data, which is extracted through subtracting the noisy output from the smoothed output. Subsequently, the binning method is combined with GPR to obtain the approximate model of the diffusion dynamics. These approximations are integrated into the SGT‐DDP framework to compute optimal control policies. Simulations on benchmark nonlinear systems under unknown dynamics demonstrate the effectiveness of the method.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1049-8923
1099-1239
DOI:10.1002/rnc.7984