Multi-task Assignment Research for Heterogeneous UAVs based on Improved Simulated Annealing Particle Swarm Optimization Algorithm

Abstract

Task assignment problem of unmanned aerial vehicles (UAVs) based on the artificial intelligent algorithms has been widely explored in recent years. UAVs’ heterogeneity including velocity, range, number of weapons is studied in this paper. Mathematical model is constructed based on the total distance objective function and complex constrains of UAVs, such as the multiple tasks, specified task sequence and time window. To solve the problem, the improved simulated annealing particle swarm optimization (SAPSO) algorithm is applied. In addition, the relationship between the particle swarm and the feasible task allocation scheme is established. The reasonable and efficient task assignment schemes are obtained based on the coding and repair- based methods. Large numbers of experimental simulations show that the improved SAPSO algorithm is more reliable and provides a reference for multi-task assignment problem of heterogenous multi-UAVs.