[Lecture] Cross-domain Multi-UAV Cooperative Task assignment problem

Speaker: Professor Wang Lei (Dalian University of Technology)

Date: December 7, 2022

Time: 16:00-18:00

Venue: Tencent Meeting (234-990-349)

Abstract: With the development of UAV technology and artificial intelligence, UAV cooperative warfare has attracted wide attention from all walks of life in recent years. In the modern battlefield environment, combat missions are becoming more and more complex in space and time, and UAVs trans-regional joint operations are becoming an overwhelming trend in modern air warfare.We aim to solve the problem of collaborative mission planning and decision-making of heterogeneous UAV groups, which takes into account the existence of multiple no-fly zones in the battlefield area. Firstly, the problem is divided by deconstruction, and many influencing factors in the cross-domain operation are considered in detail from the UAV level, target level, environment level and airport level. Then, appropriate mathematical simplification is used to reduce the complexity of the problem, and the corresponding mathematical optimization model is established for the problem of cross-domain multi-UAV cooperative mission planning. The minimization of program execution time is taken as the objective function, and complex constraints such as priority, ammunition load, obstacle avoidance and task requirements are taken into account. A two-layer algorithm is used for decoupling processing. On the upper layer, a discrete genetic algorithm is temporarily designed to solve the task allocation problem. For the lower layer, an approach based on viewable method is used to estimate the flying distance of the UAV in the dense environment of the no-fly zone. Finally, the effectiveness and scalability of the system algorithm are verified by different types of numerical simulation.