Research On Vision-based Formation Configuration Planning For Unmanned Aerial Vehicle Swarm

Different from the conventional Unmanned Aerial Vehicle(UAV)swarm formation control which requires communication between UAVs to transfer information,UAV swarm vision-based formation requires no communication but only visual measurement information to maintain formation control,which is not disturbed by complex electromagnetic environment and has an irreplaceable role in performing certain specific tasks.UAV swarm configuration is a specific arrangement composed of multiple UAVs to adapt to the environment and task requirements.Switching the swarm configuration can increase the environmental adaptability and task types of the swarm.In this paper,we investigate the planning problem of configuration switching in the context of visual formations of small swarm of UAVs:Firstly,the optimal planning problem of formation configuration switching is solved in the formation coordinate system based on the Hungarian algorithm,following the collision constraint and the through-view constraint,with the total distance of the moving path and the switching time as the optimization index.The configuration switching planning problem is decomposed into two problems: configuration switching objective assignment and path planning,and the optimization model of the objective assignment problem is established and the Hungarian algorithm is used as the solution method.The standard Hungarian algorithm is only able to solve the allocation scheme with the total distance of the mobile path as the optimization index,while the improved Hungarian algorithm,obtained by power-squared improvement of the Hungarian algorithm,is able to solve the allocation scheme with the switching time as the optimization index.At the same time,a straight-line path is used as the movement path of the UAV from the initial position to the desired position,and the allocation scheme solved with the Hungarian algorithm and the improved Hungarian algorithm is used to obtain the optimal solution of the configuration switching problem under the constraints of collision and through-view constraints,and the feasibility of the method is verified by simulation under the formation coordinate system.Secondly,for practical engineering applications,the planning problem of formation configuration switching is solved in the formation coordinate system with the optimization objective of switching time,following the collision constraint and the through-view constraint,using the orthogonal decomposition fold path and the improved power-squared Hungarian algorithm.By orthogonal decomposition translation of the straight path to obtain the fold path and establish a mathematical model,the UAV moves along the direction parallel to the coordinate axis,which is beneficial to the control analysis of the whole formation system,in addition to preventing collision to a certain extent.The solution of the configuration switching target assignment scheme under the fold path is solved by an improved power-squared Hungarian algorithm with a fold moving path to obtain the solution of the planning problem applicable to engineering applications under the constraints of collision constraints and through-view constraints,and the feasibility of the method is verified by simulation in formation coordinates.Finally,the solution results of the above two methods in the formation coordinate system are simulated in the ground coordinate system to further verify the feasibility of the methods,and the algorithm is applied to the configuration switching flight test of a four-aircraft formation to verify the practicality of the configuration switching planning method based on the orthogonal decomposition fold path in engineering applications.