Cooperative and Distributed Reinforcement Learning of Drones for Field Coverage

TL;DR

This paper introduces a distributed MARL algorithm enabling UAV teams to cooperatively cover unknown fields efficiently, addressing joint-action complexity with game theory and high-dimensional states with function approximation, validated through simulations and real-world tests.

Contribution

It presents a novel MARL approach combining game-theoretic correlated equilibrium and function approximation for UAV coverage tasks, with experimental validation.

Findings

UAV teams successfully learn to cover fields with minimal overlap.

The algorithm performs well in both simulation and physical experiments.

Cooperative learning improves coverage efficiency and reduces redundancy.

Abstract

This paper proposes a distributed Multi-Agent Reinforcement Learning (MARL) algorithm for a team of Unmanned Aerial Vehicles (UAVs). The proposed MARL algorithm allows UAVs to learn cooperatively to provide a full coverage of an unknown field of interest while minimizing the overlapping sections among their field of views. Two challenges in MARL for such a system are discussed in the paper: firstly, the complex dynamic of the joint-actions of the UAV team, that will be solved using game-theoretic correlated equilibrium, and secondly, the challenge in huge dimensional state space representation will be tackled with efficient function approximation techniques. We also provide our experimental results in detail with both simulation and physical implementation to show that the UAV team can successfully learn to accomplish the task.