Route Planning Algorithm And Application In Aircraft Route Based On GPU

In the ever-changing era of science and technology,the development speed of aircraft is very rapid.Regardless of whether it is in civil or military fields,it has always been the pursuit of intelligent unmanned development.However,the air environment is more and more complex,and there are more and more factors affecting flight.Therefore,accurate planning environment modeling and improving the efficiency of route planning algorithms are the key to improve the utilization of air routes and ensure the safe,reliable and efficient flight of aircraft.This paper first briefly introduces the application background,development status and key technologies of route planning.Then,the threat factors affecting the route planning are analyzed and modeled respectively,and a method of calculating the threat body model is proposed to obtain the waypoint quickly.Combining the related knowledge of graph theory,we turn the waypoint into the weighted adjacency matrix of the distance between storage nodes,and the matrix is the initial conditions for the study of the route planning algorithm,which is the foundation for simplifying route optimization.In this paper,the Floyd algorithm in the shortest path algorithm is selected as the route planning algorithm under the conditions of route planning and the application of different algorithms.Simulation experiments are performed to verify the accuracy and effectiveness of the algorithm.For the route planning system,when the number of aircraft and the number of threat zones reach a certain scale,the efficiency of a simple serial program cannot meet the demand for high efficiency.In order to reduce the time delay caused by the increasing scale of multi-source threat distribution in planning tasks,in this paper the CUDA architecture implements parallelism of the Floyd algorithm on the GPU.By setting up the iterative calculations in the adjacency matrix for each thread in the GPU,serial parallelism simulation experiments are performed for different scales ofplanning space.The experimental data shows that the computation speed on the GPU is significantly higher than it on the CPU,and when the calculation scale increases,the acceleration ratio continues to increase within a certain range.In order to further improve the efficiency of route planning,this paper combines OpenMP shared-memory parallel architecture to achieve multi-core CPU parallelism.Through each CPU core calling a GPU card,Floyd algorithm parallels in GPUs environment.The experimental results show that when the scale of data is same,the more GPUs,the GPU does not show greater speedup.With a certain scale,the single-GPU paralleled Floyd algorithm is more suitable as a solution for high-performance aircraft route applications.