Genetic Algorithm-Based Unmanned Aerial Vehicle (UAV) Path Planning in Dynamic Environments for Disaster Management

Abstract

Unmanned Aerial Vehicles (UAVs) hold immense potential in disaster management by enabling rapid response, real-time aerial reconnaissance, and improved situational awareness without endangering human lives. This research proposes a real-time UAV path-planning system based on a Hierarchical Recursive Multiagent Genetic Algorithm (HR-MAGA). Unlike traditional methods that struggle with adaptability in dynamic 3D environments, our system employs localized waypoint updates to reduce the computational cost of full-path recalculations. A multi-objective fitness function guides the optimization process by balancing safety, energy efficiency, altitude smoothness, turbulence resistance, and travel time. Additionally, the system integrates a decoupled real-time collision avoidance module for immediate response to sudden threats. While obstacle detection is abstracted in this study, the framework is designed to be easily integrated with real-time sensing technologies such as LiDAR for dynamic obstacle awareness. Experimental evaluations show a 20-30% improvement in path efficiency and a 40% increase in convergence speed compared to conventional genetic algorithms, highlighting the system's adaptability and robustness in disaster response scenarios.