Research On The Key Problems Of UAV Cooperative Formation

Faced with a highly confrontational,highly uncertain,and highly dynamic mission environment,the mission requirements of UAVs have gradually evolved from single-platform to multi-platform.In recent years,the technology of multi-UAV formation has developed rapidly.Hundreds and thousands of small rotorcrafts have been collaboratively performed shows many times,whereas the collaborative formation of traditional fixed-wing UAVs are yet to be developed.As one of the most mainstream configurations in the aircraft,the fixed-wing UAV can achieve greater flight speed and load capacity,but it can not perform functions such as vertical take-off and landing,hovering,etc.The flight test for fixed-wing UAVs also require a larger space.Therefore,this thesis focuses on fixed-wing UAVs.According to their characteristics,we carry out research on the key technologies of fixed-wing UAV cooperative formation.The thesis first discusses the rendzvous of multiple fixed-wing UAVs.As a pre-process of formation flight,the rendzvous task not only requires the UAV to reach the consistency of the assembly time,but also requires them to be consistent in speed and heading angle.In order to estimate the flight time of the UAV,the dynamic model of the fixed-wing UAV is introduced in detail in Chapter 2,and the simplified characteristics of the UAV are summarized according to simulation and flight test data of the previous model.We discuss the rendezvous task in two cases:The rendezvous with variable rendezvous positions.A solution based on Nash protocol is proposed for this situation.Because the formation and rendezvous location has a great influence on the rendezvous time and flight path,the 2nd chapter proposes a distributed negotiation method,which enables each drone to negotiate based on the Nash protocol method,and obtain the sub-optimal assembly position where all UAVs can fly.Then,according to the specified rendezvous position,the flying path is derived for the UAV through the Bezier spline curve.The Bezier curve is smooth everywhere,which guarantees the continuity of the path.Limiting the curvature according to the minimum turning radius of the UAV ensures that the path is feasible.Although this method also needs iterative calculation,the calculation amount is much smaller than the swarm intelligence-based optimization method,and it can still be calculated in real time.The rendezvous with specified rendezvous positions.The third chapter proposes a UAV formation rendezvous strategy based on path adjustment and speed adjustment.The rendezvous time is determined according to the shortest estimation flight time of the UAVs.A construction method of time-varying vector field is proposed.The idea of horizontal maneuver is combined with the time-varying vector field.The length of the trajectory is adjusted in real time through the time-varying vector field according to the real-time situation,and the speed is adjusted at the same time.Therefore the proposed method enables each UAV to reach the rendevous position simultaneously,and the speed and the heading angle are consistent,which meets our requirements.The convergence of the system is proved,and the handling of the collision avoidance problem in the actual situation is analyzed.The simulation results show that the proposed method has good practicability.Then,the problem of multi-UAV formation reconstruction is discussed.In addition to the terminal state constraint,time of reconfiguration,and control consumption,the constraints of control input and safety distance to avoid colliding are also considered.The reconfiguration problem is formulated as a constrained optimization problem.However,the UAV states are highly coupled and the problem is high-dimensional in this problem.We transform it to two simpler optimization problems,therefore each UAV can deal with its distributed optimization problem easily given the shared information.Two convex optimization methods,namely the Gauss-Seidel algorithm and the interior point algorithm,are applied to solve the problems.The efficiency of the proposed method is demonstrated by the simulation example results.Finally,a typical application of multiple UAV collaboration is studied: tracking ground moving targets.In the fifth chapter,a multi-UAV tracking control algorithm based on target motion state estimation is proposed.Firstly,a mathematical model is established for the problem.Then the relative distance between the target and each UAV,and the relative angles between UAVs are used as collaborative variables to achieve the coordination in task level.Each UAV calculates its control input by the backstepping method according to its expected states and the estimated target states.The extended Kalman filter method is used to estimate the target state.The related information of the UAV tracking target process is regarded as observation measurement to improve the estimation accuracy.The effectiveness of the proposed method is verified by simulation.