mlsanityData to 14 Oct 2026

A Control-Barrier-Function-Based Algorithm for Policy Adaptation in Reinforcement Learning

arXiv 2510.02720 ↗First posted 3 Oct 2025eess.SY cs.SYcited by 0

On the map: Deep reinforcement learning methods → Safe and constrained RL

Abstract (arXiv)

This paper considers the problem of adapting a predesigned policy, represented by a parameterized function class, from a solution that minimizes a given original cost function to a trade-off solution between minimizing the original objective and an additional cost function. The problem is formulated as a constrained optimization problem, where deviations from the optimal value of the original cost are explicitly constrained. To solve it, we develop a closed-loop system that governs the evolution of the policy parameters, with a closed-loop controller designed to adjust the additional cost gradient to ensure the satisfaction of the constraint. The resulting closed-loop system, termed control-barrier-function-based policy adaptation, exploits the set-invariance property of control barrier functions to guarantee constraint satisfaction. The effectiveness of the proposed method is demonstrated through numerical experiments on the Cartpole and Lunar Lander benchmarks from OpenAI Gym, as well as a quadruped robot, thereby illustrating both its practicality and potential for real-world policy adaptation.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Safe and constrained RL, one of 15 directions in Deep reinforcement learning methods. Builds on 21 of its 44 references that are in the map. First posted on arXiv 3 Oct 2025.

Builds on

21 of 44 references.

References · 21

  1. 2015Continuous control with deep reinforcement learningcited by 3,407
  2. 2016Progressive Neural Networkscited by 1,361
  3. 2019Learning agile and dynamic motor skills for legged robotscited by 504
  4. 2019Control Barrier Functions: Theory and Applicationscited by 372
  5. 2016Successor Features for Transfer in Reinforcement Learningcited by 340
  6. 2017Overcoming Exploration in Reinforcement Learning with Demonstrationscited by 340
  7. 2020Transfer Learning in Deep Reinforcement Learning: A Surveycited by 265
  8. 2022Legged Locomotion in Challenging Terrains using Egocentric Visioncited by 151
  9. 2016Learning Modular Neural Network Policies for Multi-Task and Multi-Robot Transfercited by 127
  10. 2017Learning Invariant Feature Spaces to Transfer Skills with Reinforcement Learningcited by 85
  11. 2018Dexterous Manipulation with Deep Reinforcement Learning: Efficient, General, and Low-Costcited by 84
  12. 2019Distilling Policy Distillationcited by 84
9 more
  1. 2019Pontryagin Differentiable Programming: An End-to-End Learning and Control Frameworkcited by 52
  2. 2022Safe Reinforcement Learning Using Black-Box Reachability Analysiscited by 26
  3. 2022Toward Equation of Motion for Deep Neural Networks: Continuous-time Gradient Descent and Discretization Error Analysiscited by 13
  4. 2021Optimization Algorithms as Robust Feedback Controllerscited by 9
  5. 2022Control Barrier Function Based Design of Gradient Flows for Constrained Nonlinear Programmingcited by 8
  6. 2020Learning from Human Directional Correctionscited by 6
  7. 2024Online Control-Informed Learningcited by 4
  8. 2025Constrained Optimization From a Control Perspective via Feedback Linearizationcited by 4
  9. 2022Cooperative Tuning of Multi-Agent Optimal Control Systemscited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024Constrained Reinforcement Learning with Smoothed Log Barrier Function
  2. 2024Learning Control Barrier Functions and their application in Reinforcement Learning: A Survey
  3. 2025A Review On Safe Reinforcement Learning Using Lyapunov and Barrier Functions
  4. 2022Offline Reinforcement Learning with Closed-Form Policy Improvement Operators
  5. 2024Exterior Penalty Policy Optimization with Penalty Metric Network under Constraints
  6. 2025Proactive Constrained Policy Optimization with Preemptive Penalty
  7. 2025Predictive Lagrangian Optimization for Constrained Reinforcement Learning
  8. 2024Linear Function Approximation as a Computationally Efficient Method to Solve Classical Reinforcement Learning Challenges