Research On Technologies Of Path Planning And Formation Control For Multi-UAVs Formation

With the increasing complication of the battlefield and the mission requirements in modern warfare,a single UAV mostly cannot finish a task very well.However,multi-UAVs formation flight has the unique advantage in extending the range of UAVs application and improving the total operational utility,and it is being one of the key technologies of UAV development in the future.Therefore,how to realize formation flight for multiple UAVs and take full advantage of multi-UAVs cooperation has become a challenging problem with significant theoretical value and great application prospects.Multi-UAVs formation flight is a new multi-disciplinary area of UAV study and has many technical problems remain to be solved.This dissertation focus on the problems of formation path planning,formation control,formation reconfiguration and task assignment with man in loop,during the formation flight of multiple UAVs.The main research work and innovations of this dissertation are given as follows:(1)By applying the virtual structure approach,the system model for multi-UAVs formation flight is developed.The velocity constraints and turning rate constraints and maximum climbing angle constraints of UAVs are taken into account.The maximum curvature constraints and maximum torsion constraints of formation path while the formation moving as a whole are derived,so that the planned formation path is flyable.(2)For the problem of cooperative path planning for multi-UAVs formation rendezvous,Pythagorean Hodograph(PH)curve which has curvature continuity and rational intrinsic properties is adopted.And based on co-evolutionary theory and particle swarm optimization,the cooperative co-particle swarm optimization(CCPSO)algorithm is proposed.The simultaneous arrival and collision avoidance between UAVs are achieved and the paths planned are flyable.The validity of the proposed algorithm is verified by numerical simulations in 2-D and 3-D environments.(3)A path planning method for generating a flyable path for UAVs formation flying with a given configuration is proposed,based on Delaunay map,Warshall-Floyd algorithm and the multi-population hybrid particle swarm genetic algorithm.And multiple PH curves are applied for meeting the kinematic constraints of UAVs formation.The formation path planned can avoid the threats in environment well and it is a flyable optimal path for formation flight.The validity of the proposed algorithm is verified by numerical simulations in 2-D and 3-D environments.(4)To solve the problem of the bad capacity of formation keeping,while UAVs formation executing a sharp maneuver,a formation feedback control method is studied.A nonlinear controller for formation keeping based on Lyapunov theory is designed.And a trajectory tracker with formation feedback is designed,by applying nonlinear model predictive control(NMPC)approach.The formation-keeping error can be reduced evidently during a sharp maneuver.Numerical simulations have been conducted to verify the validity of the proposed method.(5)To solve the problem of formation keeping and collision avoidance autonomously when UAVs formation flying in 3-D dynamic environment,a distributed NMPC formation controller is designed.The designed formation controllers can make UAVs steer clear of appeared static and moving obstacles with no collisions between UAVs,and the formation keeps following the reference path within the original shape after finishing obstacle avoidance.Numerical simulations in 3-D dynamic environment have been conducted to verify the validity of the proposed method.(6)To guarantee the viability and operational utility of UAVs formation in the dynamic battlefield,a rapid formation reconfiguration approach with man in loop is proposed,based on integer linear program(ILP).The approach proposed can deal with different emergency situations effectively.The validity and real-time requirement of the proposed algorithm are verified by numerical simulations.(7)To solve the problem of task reassignment for schooling UAVs attacking multiple targets within multi-formations in dynamic battlefield,the online task reassignment method with man in loop for multiple formations is studied,which is based on the Delaunay map and the improved consensus-based auction approach(CBAA).So the multi-formations can adjust the original task assignment rapidly when the battlefield situation changes.Simulations under different emergency situations are conducted to verify the method effectiveness and real-time performance.