Research And Implementation Of A General Platform For Cooperative Task Planning With Multi-UAVs

In recent years,the rapid development of Unmanned Aerial Vehicle(UAV)technologies has made it capable to perform various complex tasks,and enables various applications such as transportation,geological survey,information collection,and disaster rescue.In order to overcome the limitations of single UAVs and improve the efficiency of task completion,multi-UAV collaboration has become a research hotspot.In realistic applications,however,we are facing with several important issues such as diversified UAV types,poor cross-platform collaboration,and lack of effective path planning algorithms.Therefore,it is of great research significance and with practical value to build a general platform for cooperative task planning,and solve the cross-platform UAV joint control problem and multi-UAV path planning problem.In order to build a general platform for cooperative task planning with multi-UAVs,this thesis first designs a hierarchical cross-platform joint control architecture,including the control layer,communication layer,and application layer.By using the underlying module to shield the differences of UAVs,the upper layer can operate multi-UAVs of different models in the same control system by treat them as same.Based on this architecture,we respectively design and implement the cross-platform UAV control module,communication module and ground station software.Second,we propose an online path planning algorithm based on the improved variable neighborhood search scheme for solving the multi-UAVs path planning problem in a dynamic environment,which can achieve two optimization objectives,i.e.,minimizing the total energy consumption of multi-UAVs and minimizing the maximum energy consumption of multi-UAVs.Finally,based on the Pixhawk and DJI M100 flight control platforms,we build a general platform prototype for cooperative task planning with multi-UAVs.Meanwhile,considering a realistic application requirement of constructing panoramic maps using multi-UAVs,we conduct extensive performance evaluations on various modules of the platform and the proposed path planning algorithm.The results verify the feasibility and effectiveness of the platform in real-world applications.