Research On Route Planning Of Quadrotor UAV Based On Improved A~* Fusion Algorithm

Quadrotor unmanned aerial vehicles(UAV)has the characteristics simple structure and high mobility.It can not only take off vertically,but also quickly change the direction of flight in the air.The quadcopter is easy to operate and portable,which has been wildly used in civil and military.It has been a hot research topic both domestically and internationally.Quadrotor UAV is a typical complex system which is underactuated,nonlinear and strongly coupled.With the progress of times and the development of science and technology,higher requirements have been placed on the adaptability of quadrotor UAV,which need to accomplish tasks in more complex environments and reduce the risk of personnel involvement.Therefore,route planning in the process of flight is particularly important.This thesis mainly studies the route planning of plant protection UAVs,whose main research content is as follows:(1)The structure,flight principle and dynamics of the quadrotor are analyzed.Based on the Newton-Euler equation,the rigid body model of the UAV is established.On the basic of the kinematics equation of UAV,a route planner using A~* algorithm is designed.Firstly,A~* algorithm quickly generates a series of discrete waypoints in the grid map.Then,considering the kinematics of the UAV,trajectory optimization algorithm based on minimum snap is used to smooth the route.Simulation results that the route optimized by A~* can better adapt to the flight of UAV,avoiding the pause due to the adjustment of heading.The route is smooth,stable and safe,and the flight time is also saved.(2)The improved ant colony algorithm is used to search the path of hill map.The evaluation function of A~* algorithm is used to improve the heuristic function of the traditional ant colony algorithm(ACO),and the UAV height limiting factor is added.Then,adaptive rules are added to the pheromone updating and accumulation process.The ants on the optimal path in each iteration are selected by elite strategy.The simulation shows that the improved ant colony algorithm can search the optimal path quickly without falling into the local optimal compared with the traditional ant colony algorithm.At the same time,all waypoints have safe distance above the hilly surface,suitable for the passage of UAV.(3)Compared with the traditional A~* algorithm,Hybrid A~* algorithm with two kinds of heuristic functions and search trees can find the optimal route quickly and accurately.The simulation shows that Hybrid A~* algorithm has high search efficiency,but in the map with obstacles,some routes are close to obstacles,which does not meet the flight safety requirements of the quadrotor UAV.Therefore,based on the lidar carried by the UAV and the local planning algorithm,VFH,the simulation results that it can predict and avoid obstacles in advance compared with Hybrid A~* algorithm and better adapt to the flight of the quadrotor UAV.In summary,based on the classical A~* algorithm,the trajectory optimization algorithm minimum snap is used to smooth the waypoints generated by A~* firstly.Then,the heuristic function of A~* which considers the position of initial node and target point is used to improve the ant colony algorithm.Finally,based on the original single heuristic function of A~*,the Hybrid A~* algorithm is used to further improve the search speed,and the local route planning algorithm VFH algorithm improves the flight safety of UAV.At the end of the thesis,the global and local route planner is applied to the plant protection UAV.Simulation proves that the improved UAV can complete the task in the static environment with obstacles.