A Novel Efficient Particle Swarm Optimization Algorithm for UAV Applications in Bushfire Hazard Management

Abstract

Owing to the drastic environmental change and frequently occurring disasters like bushfires in Australia and New South Wales (NSW) in particular, this research study aims to develop an unmanned aerial vehicles (UAVs) assisted bushfire detection and prevention system. A rule-based verification framework is developed to identify critical infrastructure impacted by bushfires using a metaheuristic optimization algorithm, with reliance on real-time data. Moreover, the affected area coverage of UAVs is maximized by considering limited resources and energy constraints for better precision of results. Finally, a mathematical optimization model is developed and tested for path planning of different UAV populations in affected bushfire regions. Within the realm of competing techniques for UAVs path planning, this research study extensively scrutinized and assessed four distinct optimization algorithms, including whale optimization algorithm (WOA), grey wolf optimizer (GWO), and particle swarm optimization (PSO). Results indicated that our proposed model, efficient particle swarm optimization (EPSO), consistently outperforms other techniques across all cases. However, the degree of its superiority varies among different scenarios. In Case 3, EPSO exhibits the most dominance, presenting notable improvements with approximately 18.35% to 54.47% reductions in cost and time compared with WOA, GWO, and PSO.