Research And Simulation Of Intelligent Formation Algorithm For Multi-rotor Unmanned Aerial Vehicle

The multi-rotor UAV has the characteristics of simple control,low cost,low requirements for taking off and landing,and freely hovering in flight.Transforming the multi-rotor UAV into the intelligent agent to realize intelligent formation,can greatly enhance the application value of the multi-rotor UAVs group.The flight control principle of multi-rotor UAV is introduced,and the intelligent formation system of multi-rotor UAV is designed.The system adopts a multi-agent framework,and the drone acquires information to complete the action selection independently.Through the flight controller,the instructions of the action are converted into the movement of the drone to realize intelligent formation.Firstly,the flow field of multi-rotor UAV is analyzed by the combination of simulation and experiment.The characteristics of multi-rotor UAV flow field under different motion states are analyzed,and the airflow interference area of the dome is determined.It is proposed that the multi-rotor UAV space formation should use a layered formation.Secondly,the target airspace point autonomous selection algorithm is studied.The sparse formation uses the Hungarian algorithm with improved parameters to optimize the average flight distance and formation completion time;the dense formation uses the Hungarian algorithm and the improved Hungarian algorithm to achieve dense formation and formation complementation.Then two kinds of action decision methods based on position and distance are proposed.The position-based action decision method uses the membership function to synthesize the action of the UAV and the target airspace point position,velocity and flow field.The distance-based action decision method needs to install the distance sensor.At the same time,the vertical movement is separated from the horizontal movement when the formation is formed,and finally the intelligent formation is realized.Then the hardware used in the article was tested to verify the functions of the multirotor UAV,flight controller and onboard computer,ultrasonic sensor and communication module.Finally,combined with the full-text research content,the position-based and distance-based action decision-making methods are simulated and verified.With the target air-space point selection algorithm,the multi-rotor UAV group can smoothly and efficiently realize the formation and formation transformation,and at the same time integrate the impact of airflow interference in the formation process.