Adaptive Primal-Dual Method for Safe Reinforcement Learning

TL;DR

This paper introduces an adaptive primal-dual method for safe reinforcement learning that dynamically adjusts learning rates, leading to improved stability and performance in constrained policy optimization tasks.

Contribution

The paper proposes a novel adaptive primal-dual algorithm with theoretical guarantees and demonstrates its effectiveness over existing methods in various environments.

Findings

Outperforms constant LR methods in stability and performance

Achieves comparable or better results than state-of-the-art SRL algorithms

Demonstrates robustness of adaptive learning rates through empirical evidence

Abstract

Primal-dual methods have a natural application in Safe Reinforcement Learning (SRL), posed as a constrained policy optimization problem. In practice however, applying primal-dual methods to SRL is challenging, due to the inter-dependency of the learning rate (LR) and Lagrangian multipliers (dual variables) each time an embedded unconstrained RL problem is solved. In this paper, we propose, analyze and evaluate adaptive primal-dual (APD) methods for SRL, where two adaptive LRs are adjusted to the Lagrangian multipliers so as to optimize the policy in each iteration. We theoretically establish the convergence, optimality and feasibility of the APD algorithm. Finally, we conduct numerical evaluation of the practical APD algorithm with four well-known environments in Bullet-Safey-Gym employing two state-of-the-art SRL algorithms: PPO-Lagrangian and DDPG-Lagrangian. All experiments show that the practical APD algorithm outperforms (or achieves comparable performance) and attains more stable training than the constant LR cases. Additionally, we substantiate the robustness of selecting the two adaptive LRs by empirical evidence.