Design And Realization Of Flight Control For UAV Group's Autonomous Formation

UAV(Unmanned Aerial Vehicle)formation can complete multiple tasks and improve the efficiency of task completion at the same time,which can be realized by self-organizing control.UAV formation is an important direction for the development of UAVs at present,which can be applied to multiple scenarios such as search and rescue,battlefield reconnaissance,any military strike.It has been widely concerned because of its advantages of expanding the scope of tasks,improving the efficiency of task execution,increasing the quality of task completion,improving the system's fault tolerance,and expanding task capabilities.In this paper,several quadcopters have been designed and produced to form a UAV formation.To achieve UAV formation flight,the quadcopter in the formation must be able to perceive its position in space.The DWM1000 module was selected as the ultra-wideband signal transceiver module,and combined with the STM32F103C8T6 microcontroller to design the indoor positioning software and hardware system.The system includes anchors and tags.The four anchors are fixed at different positions,and the coordinates are marked.The tags are carried on the quadcopter to detect position.The hardware and software of the wireless communication system are based on the NRF24L01 wireless communication module.The wireless communication system is designed to realize the communication between the quadcopter and the host computer,including the USB adapter as a communication bridge between the drone and the host computer.The frequency-hopping technology is used to realize one-to-many communication.The distance between the label and the four anchors is obtained by Two Way Ranging(TWR).The data is pre-processed by limiting filtering and the Kalman filtering,and the measurement data is corrected by first-order linear fitting,which improves the measurement accuracy.A cooperative positioning algorithm based on trilateration and the full centroid algorithm,which solves the spatial coordinates of the label,and gives the positioning error rule of each axis in the solution result,and uses this as the estimation error to obtain the final positioning result of each axis.The static and dynamic positioning experiments show that this method is effective.The formation mode of virtual leader-follower is adopted,and three formation formations are designed.The artificial potential field method based on judging the local minimum point and escaping the local minimum point is designed to plan the virtual leader path.Use the path optimization algorithm of the regression search method to optimize the path globally.Since the optical flow sensor carried on the UAV has poor data coincidence on the z-axis of the UAV,the three-dimensional path planning problem is reduced to the two-dimensional path planning problem.The host computer is designed and implemented in C# programming language,and the feasibility of the system is verified through experiments.