A hybrid algorithm of UAV path planning for rescue in bushfire environments

Abstract

This paper presents a novel hybrid algorithm aimed at optimizing the planning of forest fire rescue routes. The proposed method utilizes a hierarchical architecture to ensure safe navigation deployment, particularly in environments that are either unknown or only partially known. The observation layer effectively deploys a comprehensive set of feasible navigation points through the utilization of global path planning techniques. Subsequently, the execution layer takes charge of executing these navigational actions. To further enhance safety, the decision layer assesses whether the unmanned aerial vehicle (UAV) requires local obstacle avoidance strategies. Furthermore, an additional decision regarding replanning is incorporated into the decision layer, addressing the potential risks associated with dynamic avoidance approaches. This consideration effectively mitigates issues like being trapped in a perpetual loop or encountering path-finding challenges.