An efficient method for Multiple UAV Cooperative Forest Fire Reconnaissance Task Assignment
Xinlin Liu, Jing Tian, Linyi Hou · Research Square · 2022
Abstract In this paper, we present a three-layer centralized multiple unmanned aerial vehicles(UAVs) task assignment method in collaboration with satellites for forest fire reconnaissance task, based on Gaussian mixture model (GMM) and multidimensional 0-1 knapsack model. The goal of UAVs reconnaissance task assignment problem is to maximize the utilization of UAVs by the reasonable allocation of UAV group. Hence, it is taken as a combinatorial optimization problem in this paper. We first obtain the basic information of the fire field through the space-based infrared sensors carried by satellites. Then we fit the fire points distribution through expectation maximization algorithm (EM) based on GMM to obtain the initial position of each UAV. Finally, we consider the task assignment problem of UAVs as a multidimensional 0-1 knapsack problem and use the improved genetic algorithm (GA) inspired by the fireworks algorithm (FWA) with improved select operator and elite opposition-based learning strategy (SeEl-FWGA) to solve the UAVs task assignment problem. Finally, the proposed method is demonstrated and compared with other methods under the background of forest fires in Liangshan Prefecture, Sichuan Province. The simulation results show the superiority to the traditional genetic algorithm in terms of iteration speed and iteration results, and the high efficiency to solve the UAVs task assignment problem.