Saved in:
Bibliographic Details
Main Authors: Yaghmaie, Farnaz Adib, Naha, Arunava
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2511.11131
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908653091553280
author Yaghmaie, Farnaz Adib
Naha, Arunava
author_facet Yaghmaie, Farnaz Adib
Naha, Arunava
contents Flow $Q$-learning has recently been introduced to integrate learning from expert demonstrations into an actor-critic structure. Central to this innovation is the ``the one-step policy'' network, which is optimized through a $Q$-function that is regularized with the behavioral cloning from expert trajectories, allowing learning more expressive policies using flow-based generative models. In this paper, we studied the convergence property and stabilizablity of the one-step policy during learning for linear quadratic problems under the offline settings. Our theoretical results are based on a new formulation of the one-step policy loss based on the average expected cost, and regularized with the behavioral cloning loss. Such a formulation allows us to tap into existing strong theoretical results from the policy gradient theorem to study the convergence properties of the one-step policy. We verify our theoretical finding with simulation results on a linearized inverted pendulum.
format Preprint
id arxiv_https___arxiv_org_abs_2511_11131
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Convergence of Flow-Policy Gradient Learning for Linear Quadratic Regulator Problems
Yaghmaie, Farnaz Adib
Naha, Arunava
Systems and Control
Flow $Q$-learning has recently been introduced to integrate learning from expert demonstrations into an actor-critic structure. Central to this innovation is the ``the one-step policy'' network, which is optimized through a $Q$-function that is regularized with the behavioral cloning from expert trajectories, allowing learning more expressive policies using flow-based generative models. In this paper, we studied the convergence property and stabilizablity of the one-step policy during learning for linear quadratic problems under the offline settings. Our theoretical results are based on a new formulation of the one-step policy loss based on the average expected cost, and regularized with the behavioral cloning loss. Such a formulation allows us to tap into existing strong theoretical results from the policy gradient theorem to study the convergence properties of the one-step policy. We verify our theoretical finding with simulation results on a linearized inverted pendulum.
title Convergence of Flow-Policy Gradient Learning for Linear Quadratic Regulator Problems
topic Systems and Control
url https://arxiv.org/abs/2511.11131