Attention Actor-Critic algorithm for Multi-Agent Constrained Co-operative Reinforcement Learning

TL;DR

This paper extends the Attention Actor-Critic algorithm to multi-agent constrained reinforcement learning, enabling agents to optimize shared goals while satisfying action constraints through different attention modes, leading to improved performance.

Contribution

It introduces a novel multi-mode attention mechanism within Actor-Critic for constrained multi-agent RL, enhancing the ability to optimize goals and satisfy constraints simultaneously.

Findings

Effective in benchmark environments

Improves action quality under constraints

Demonstrates better convergence behavior

Abstract

In this work, we consider the problem of computing optimal actions for Reinforcement Learning (RL) agents in a co-operative setting, where the objective is to optimize a common goal. However, in many real-life applications, in addition to optimizing the goal, the agents are required to satisfy certain constraints specified on their actions. Under this setting, the objective of the agents is to not only learn the actions that optimize the common objective but also meet the specified constraints. In recent times, the Actor-Critic algorithm with an attention mechanism has been successfully applied to obtain optimal actions for RL agents in multi-agent environments. In this work, we extend this algorithm to the constrained multi-agent RL setting. The idea here is that optimizing the common goal and satisfying the constraints may require different modes of attention. By incorporating different attention modes, the agents can select useful information required for optimizing the objective and satisfying the constraints separately, thereby yielding better actions. Through experiments on benchmark multi-agent environments, we show the effectiveness of our proposed algorithm.