Adaptive deep learning for high-dimensional Hamilton--Jacobi--Bellman equations

T Nakamura-Zimmerer, Q Gong, W Kang - SIAM Journal on Scientific …, 2021 - SIAM
Computing optimal feedback controls for nonlinear systems generally requires solving
Hamilton--Jacobi--Bellman (HJB) equations, which are notoriously difficult when the state …

Overcoming the curse of dimensionality for some Hamilton–Jacobi partial differential equations via neural network architectures

J Darbon, GP Langlois, T Meng - Research in the Mathematical Sciences, 2020 - Springer
We propose new and original mathematical connections between Hamilton–Jacobi (HJ)
partial differential equations (PDEs) with initial data and neural network architectures …

Tensor decomposition methods for high-dimensional Hamilton--Jacobi--Bellman equations

S Dolgov, D Kalise, KK Kunisch - SIAM Journal on Scientific Computing, 2021 - SIAM
A tensor decomposition approach for the solution of high-dimensional, fully nonlinear
Hamilton--Jacobi--Bellman equations arising in optimal feedback control of nonlinear …

Using adaptive sparse grids to solve high‐dimensional dynamic models

J Brumm, S Scheidegger - Econometrica, 2017 - Wiley Online Library
We present a flexible and scalable method for computing global solutions of high‐
dimensional stochastic dynamic models. Within a time iteration or value function iteration …

Polynomial approximation of high-dimensional Hamilton--Jacobi--Bellman equations and applications to feedback control of semilinear parabolic PDEs

D Kalise, K Kunisch - SIAM Journal on Scientific Computing, 2018 - SIAM
A procedure for the numerical approximation of high-dimensional Hamilton--Jacobi--
Bellman (HJB) equations associated to optimal feedback control problems for semilinear …

On some neural network architectures that can represent viscosity solutions of certain high dimensional Hamilton–Jacobi partial differential equations

J Darbon, T Meng - Journal of Computational Physics, 2021 - Elsevier
We propose novel connections between several neural network architectures and viscosity
solutions of some Hamilton–Jacobi (HJ) partial differential equations (PDEs) whose …

Learning optimal feedback operators and their sparse polynomial approximations

K Kunisch, D Vásquez-Varas, D Walter - Journal of Machine Learning …, 2023 - jmlr.org
A learning based method for obtaining feedback laws for nonlinear optimal control problems
is proposed. The learning problem is posed such that the open loop value function is its …

Optimal feedback law recovery by gradient-augmented sparse polynomial regression

B Azmi, D Kalise, K Kunisch - Journal of Machine Learning Research, 2021 - jmlr.org
A sparse regression approach for the computation of high-dimensional optimal feedback
laws arising in deterministic nonlinear control is proposed. The approach exploits the control …

QRnet: Optimal regulator design with LQR-augmented neural networks

T Nakamura-Zimmerer, Q Gong… - IEEE Control Systems …, 2020 - ieeexplore.ieee.org
In this letter we propose a new computational method for designing optimal regulators for
high-dimensional nonlinear systems. The proposed approach leverages physics-informed …

Mitigating the curse of dimensionality: sparse grid characteristics method for optimal feedback control and HJB equations

W Kang, LC Wilcox - Computational Optimization and Applications, 2017 - Springer
We address finding the semi-global solutions to optimal feedback control and the Hamilton–
Jacobi–Bellman (HJB) equation. Using the solution of an HJB equation, a feedback optimal …