Research On Uav Patn Planning Algorithm Based On Improved Ant Colony-Dstar Hybrid Algotithm

With the progress of computer technology,communication technology and automation control technology,Unmanned Aerial Vehicles(UAVs)technology has developed rapidly in recent years.UAV has the advantages of easy deployment,mobility and flexibility.To a certain extent,UAVs can replace manned operation to accomplish tasks with higher efficiency and lower cost.UAVs path planning technology is to calculate a flight path from the source point to the target point according to the mission target,and it is the core technology of UAV mission planning.Path planning algorithms can be divided into global path planning algorithms and local real-time path planning algorithms based on whether global map information is known or not.The single global planning algorithm or the local planning algorithm always has both advantages and disadvantages.In general,static global path planning algorithms are outstanding in finding global optimal solution or approximate optimal solution,but they are unsuitable for real-time path planning in dynamic environment.Real-time dynamic path planning algorithms can interact with the environment in real time to adapt to the dynamic environment,nevertheless,they lack global optimality.Aiming at the shortcomings of the existing path planning algorithms,this paper improves the Ant Colony Optimization(ACO)and combines it with Dstar algorithm.A hybrid UAVs path planning algorithm based on improved ACO and Dstar algorithm is proposed.Firstly,this paper builds the UAV path planning space into a grid map model.The principle and implementation steps of ACO are systematically discussed,the advantages and disadvantages of ACO are analyzed as well.For the maps with known global environmental information,the defects of the classical ACO in the complex obstacle environment are verified by theoretical analysis and simulation experiments,and an improved ACO is proposed for these problems.Specifically,the initial distribution of pheromones is optimized based on direction gui dance information,which reduces the time cost of initial search.By improving the rules of pheromone volatilization and update,the ability of the algorithm to retain excellent path information is enhanced,and the convergence speed of the ant colony algorithm is accelerated.In addition,a new probabilistic transfer strategy based on regional security is proposed,which solves the problem that classical ACO is easy to fall into local deadlock,which improves the success rate of task planning.Finally,the algorithm is simulated based on maps of different scales and complexities,compared with the original algorithm,the improved ACO has significant advantages.Considering the problem of poor timeliness and adaptability of global path planning algorithm and poor global optimality of local path planning algorithm in the single path planning algorithm scheme.Based on the improved ACO and considering the dynamic environment with unexpected obstacles,a hybrid path planning scheme based on improved ACO and Dstar is proposed.Firstly,the algorithm makes full use of the advantages of swarm intelligence of ACO to output a low-cost initial path based on static environment.At the same time,Dstar algorithm is introduced to deal with sudden obstacles.In detail,once sudden obstacles were discovered,corresponding track correction can be made to avoid collision,which reconstructs the initial path only in the area covered by sudden obstacles,and there is no need to search the whole path again.Consequently,the corrected path can be obtained quickly.In the simulation part,the hybrid algorithm is compared with the improved ACO and Dstar algorithm respectively,It is verified that the hybrid algorithm can organically fuse the global path planning algorithm and local real-time path planning algorithms,which make full use of global prior information and local posterior information and finally achieves both optimization and real-time efficiency of the output path.