Safe and Energy-Aware Decentralized PDE-Constrained Optimization-Based Control of Multi-UAVs for Persistent Wildfire Suppression

TL;DR

This paper introduces a decentralized, safety- and energy-aware control framework for multi-UAV wildfire suppression, combining control Lyapunov and barrier functions, validated through simulations and real experiments.

Contribution

It develops a novel decentralized control approach that ensures safety and energy efficiency for large-scale wildfire suppression with UAVs.

Findings

Controllers effectively prevent danger zone entry.

UAVs successfully reach charging stations.

Fire suppression performance is validated in real experiments.

Abstract

This paper presents a safe and energy-aware optimization-based control framework for multi-UAV wildfire suppression under localization and motion uncertainties. We first develop a centralized density-based controller that couples UAV motion and water deployment in a wildfire-specific control Lyapunov function. This framework is then extended to a decentralized setting suitable for large-scale operations using only local information. The controllers use control barrier function constraints to enforce both danger zone avoidance and the ability to reach a charging region. Simulations and real quadcopter experiments demonstrate the controller's effectiveness in fire suppression while preserving safety and energy sufficiency over multiple charge cycles.