Abstract: A parallel A* algorithm is proposed to accomplish the aerocraft in low altitude complex geographic environment for 3D route planning. This algorithm designs a new parallel form of A* algorithm by threading its basic operations with OpenMP, ultimately, accomplishes flight path searching process to avoid obstruct in the 3D space with route constrains. Visualization simulation system is established and the models of terrain are designed and feature of scene are completed by modeling tool Creator and development tool Vega, the demonstration which validates the above algorithm. Simulation results prove that the parallel algorithm can get better search speed than serial form, the speedup and the efficiency of the algorithm increase with increasing scale of TSP.

Key words: multi-core CPU, parallel A* algorithm, OpenMP, threading

摘要： 采用一种并行的A*算法实现了复杂地理环境下低空飞行器的三维路径规划。该算法将A*算法的基本操作设计成并行执行形式，并利用OpenMP将其线程化，结合约束条件，最终完成在三维空间中避开障碍物的路径搜索过程。使用建模工具Creator和开发工具Vega为仿真平台，完成对复杂地形地貌场景的设计，在此基础上对上述算法进行了验证。仿真结果表明该并行方式较之串行方式能获得更快的寻优速度，加速比也随着规模的增大而提高。

关键词: 多核CPU, 并行A*算法, OpenMP, 线程化