Navigation for Prioritized Wildfire Boundary Tracking

This paper deals with the problem of prioritized wildfire boundary tracking with an unmanned aerial vehicle (UAV). Different from the strategy of circling the entire wildfire boundary, this paper aims to navigate the UAV to track the highly active segments of the wildfire boundary such that firefighters can adopt rapid and appropriate responses to extinguish fires on the faster propagating boundary segment. To realize prioritized tracking, an augmented area based metric is leveraged to identify the moving direction of the UAV and a vector field based algorithm is developed to navigate the UAV to converge to and patrol along the assigned wildfire segment. Computer simulations have been conducted with FARSITE simulator sysnthesized wildfire data to compare the performance of the proposed prioritized tracking strategy and a vector field based cyclic strategy. Simulation results indicates that, under the guidance of preferential tracking flights, the UAV always fly to monitor the highly active segment of the dynamic wildfire boundary.