Multiple UAVs Forest Fire Fighting using a hierarchical task planning method

A task planning approach is proposed for multiunmanned aerial vehicles (UAVs) in a forest fire fighting mission considering obstacles and simultaneous arrival. The approach consists of three stages, in the first stage, the estimate arrival time between UAVs and targets are calculated by the rapidly-exploring random tree (RRT) algorithm, which serves as input to solve task assignment problem. Then, a distributed task assignment method using auction mechanism is employed to allocate these fire spots aiming to minimize the fire-fighting mission time, the value and resources requirement of fire spots are considered. In the last stages, a modified cooperative particle swarm optimization algorithm is designed for these UAVs to reach the fire spot simultaneously, kinematic constraint of UAV and collision avoidance with mountains are considered. Simulation examples are used to validate the proposed methods.