Policy gradient method for LQG control via input-output-history representation: convergence to O()-stationary points

DOI10.1016/J.SYSCONLE.2025.106289MaRDI QIDQ6857736zbMATH OpenFDO

Authors Tomonori Sadamoto, Takashi Tanaka

Publication date 16 February 2026

Published in Systems \& Control Letters (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1016/J.SYSCONLE.2025.106289

zbMATH Keywords

linear quadratic Gaussiannon-convex optimizationpolicy gradient method

Mathematics Subject Classification ID

Nonconvex programming, global optimization (90C26) Feedback control (93B52)

Cites work

• Approximation of Large-Scale Dynamical Systems
• Convergence and Sample Complexity of Gradient Methods for the Model-Free Linear–Quadratic Regulator Problem
• Discrete-time stochastic systems. Estimation and control.
• Distributed Reinforcement Learning for Decentralized Linear Quadratic Control: A Derivative-Free Policy Optimization Approach
• scientific article; zbMATH DE number 1001723 (Why is no real title available?)
• New results in linear filtering and prediction theory
• On the optimization landscape of dynamic output feedback linear quadratic control
• On the perturbation of the Moore-Penrose inverse of a matrix
• On Topological Properties of the Set of Stabilizing Feedback Gains
• Optimizing static linear feedback: gradient method
• Policy gradient methods for designing dynamic output feedback controllers
• Structured output feedback control for linear quadratic regulator using policy gradient method

This page was built for publication: Policy gradient method for LQG control via input-output-history representation: convergence to \(\mathcal{O}(\epsilon)\)-stationary points

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6857736)