A Newton--Kleinman Iteration for Infinite-Dimensional Linear-Quadratic Optimal Control
Abstract
We develop a Newton--Kleinman iteration for linear-quadratic optimal control of infinite-dimensional systems with possibly unbounded control and observation operators. In this setting, the classical algebraic Riccati equation need not be meaningful and therefore cannot serve as the basis of the iteration. We instead linearize a Lur'e identity formulated directly on the system-node domain. Starting from any stabilizing feedback, the resulting iteration is well defined, preserves stability, and yields a monotonically decreasing sequence of cost operators. A resolvent-based normalization makes the construction applicable to unbounded control. Under finite-cost detectability and for a finite-dimensional input space, the iterates converge strongly to the optimal value operator, independently of the fixed normalization parameter. In finite dimensions, the natural zero-state normalization recovers exactly the classical Newton--Kleinman iteration. The formulation leads to a low-rank numerical realization based on shifted system solves and inversions on the input space. We illustrate the method on a linearized cylinder wake with boundary traction control and unbounded observation, using one Newton--Kleinman update per adaptively refined spatial mesh, and on a damped linear elasticity problem with boundary traction control and boundary displacement observation.
Related
- Sub-Infinite Horizon Stochastic Linear-Quadratic Optimal Control Problems and Delayed Backward Riccati Equations
- Policy Iteration for Domain Randomized Linear Quadratic Systems
- Optimal Oracle Complexity of Dual Halpern Iteration in Normed Spaces
- Quadratic polynomials preserving free infinite divisibility
- Infinite families of pairs of real quadratic fields whose class numbers are divisible by a given integer
BibTeX
@misc{reis2026newtonkleinman,
title = {{A Newton--Kleinman Iteration for Infinite-Dimensional Linear-Quadratic Optimal Control}},
author = {Timo Reis},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.03242},
eprint = {2610.03242},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.03242},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - A Newton--Kleinman Iteration for Infinite-Dimensional Linear-Quadratic Optimal Control
AU - Timo Reis
PY - 2026
DA - 2026/10/02
PB - arXiv
DO - 10.48550/arxiv.2610.03242
UR - https://arxiv.org/abs/2610.03242
AB - We develop a Newton--Kleinman iteration for linear-quadratic optimal control of infinite-dimensional systems with possibly unbounded control and observation operators. In this setting, the classical algebraic Riccati equation need not be meaningful and therefore cannot serve as the basis of the iteration. We instead linearize a Lur'e identity formulated directly on the system-node domain. Starting from any stabilizing feedback, the resulting iteration is well defined, preserves stability, and yields a monotonically decreasing sequence of cost operators. A resolvent-based normalization makes the construction applicable to unbounded control. Under finite-cost detectability and for a finite-dimensional input space, the iterates converge strongly to the optimal value operator, independently of the fixed normalization parameter. In finite dimensions, the natural zero-state normalization recovers exactly the classical Newton--Kleinman iteration. The formulation leads to a low-rank numerical realization based on shifted system solves and inversions on the input space. We illustrate the method on a linearized cylinder wake with boundary traction control and unbounded observation, using one Newton--Kleinman update per adaptively refined spatial mesh, and on a damped linear elasticity problem with boundary traction control and boundary displacement observation.
ER -