With the rapid development of science and technology,Unmanned Aerial Vehicles(UAVs)play an increasingly important role in the military and civilian fields,and they are gradually developing in the direction of cluster collaboration.As an important development direction of drone cluster collaboration,multi-UAV collaborative formation technology is still in the stage of exploration and development,and there are still many theoretical and technical issues that need to be resolved urgently.Common formation algorithms such as consensus theory,artificial potential field method and model predictive control still have real-time,reliability,robustness,and effectiveness of obstacle avoidance methods.Therefore,in response to the above problems,this thesis mainly studies the problems existing in the formation and maintenance of the UAV group,formation reconstruction,and collision avoidance and obstacle avoidance.The main research work and results are as follows:1)According to the formation process of UAV formation and based on the polynomial trajectory optimization algorithm,this thesis proposes a distributed formation trajectory planning algorithm suitable for formation and maintenance of the formation,and verifies the convergence of the algorithm.The algorithm enables each formation member to calculate their own formation trajectory based on the trajectory information of their neighboring UAVs.After the algorithm converges,the trajectory can be tracked to maintain the formation without realtime communication,which is robust.In addition,the process of solving the trajectory of the algorithm can be transformed into a quadratic programming problem,ensuring the real-time performance of the algorithm.Finally,this thesis linearizes the nonlinear UAV model based on the Lie derivative,obtains the linearized UAV constraints,and combines them with the distributed trajectory optimization algorithm to ensure the reliability of the trajectory.The above characteristics make the algorithm more feasible in practical applications.2)In terms of formation reconstruction,this thesis firstly based on the formation description method of,and gave the formation description process based on the least square method under the framework of formation trajectory planning algorithm.And for formation switching,the research content is divided into target allocation and path planning,and a target allocation algorithm based on the optimal distance is given.For formation scaling,this thesis combines an algorithm that scales formation proportionally with time and a distributed formation trajectory optimization algorithm.Finally,the distributed formation trajectory planning algorithm proposed in this thesis has the functions of formation switching and formation zooming,and the time period of formation zooming can be customized,which improves the function of the algorithm.3)For the problem of collision avoidance and obstacle avoidance in airspace during formation flight,this thesis proposes an obstacle avoidance strategy.With sufficient time,the obstacle avoidance trajectory of formation is re-planned at the level of trajectory planning.When it is not enough to re-plan the trajectory,rely on the obstacle avoidance controller at the trajectory tracking level to avoid obstacles.At the level of trajectory planning,this thesis transforms the obstacle avoidance method of Buffered Voronoi Cell(BVC)into a linear constraint,and combines it with the formation trajectory planning algorithm.The linear obstacle avoidance hard constraint not only guarantees the real-time performance of the algorithm,but also ensures the reliability of the algorithm.At the trajectory tracking level,according to the collision principle of the collision cone,this thesis improves the repulsive force generation radius of the artificial potential field method,and combines it with the PD trajectory tracking controller according to the UAV obstacle avoidance process,and proposes a response to pop-up The threat trajectory tracking controller can make the UAV continue to track a given trajectory after avoiding obstacles,ensuring the real-time and effectiveness of the obstacleavoidance algorithm. |