地标导向启发式路径规划算法.docVIP

地标导向的启发式路径规划算法 摘要:为提高大规模交通网络路径规划算法的查询效率，以a*算法为基础，提出一种地标导向的启发式算法。在预处理中将重要的顶点和边选为地标，在点对点寻径时使用地标作为启发式函数的启发参数，并进行分段计算。实验结果表明，此算法在处理长距离的路径规划问题时有较高的查询效率和更合理的计算结果。 关键词:路径规划; 地标; 预处理; 层次缩减算法; 三角启发算法  landmark.oriented heuristic routing algorithm in traffic network  meng ke*, zhang chun.yan school of computer science and technology, china university of mining and technology, xuzhou jiangsu 221008, china abstract: to improve the query efficiency of road routing algorithm in large-scale traffic network, a landmark-oriented algorithm based on a* algorithm was proposed. select the most important vertexes and edges as landmarks during preprocessing, choose appropriate landmarks as the reference parameters and calculate in sections in point-to-point routing. experiment results indicate that it has higher query efficiency and more reasonable results in long-distance road routing. to improve the query efficiency of road routing algorithm in large.scale traffic network, a landmark.oriented algorithm based on a* algorithm was proposed. select the most important vertexes and edges as landmarks during preprocessing, choose appropriate landmarks as the reference parameters and calculate in sections in point.to.point routing. the experimental results indicate that it has higher query efficiency and more reasonable results in long.distance road routing. key words: path.planning; landmark; preprocessing; contraction hierarchies (ch) algorithm; a* landmarks triangle (alt) algorithm 0 引言 对于大规模交通网络，dijkstra算法［1］需要花费长时间进行计算，不符合实时性的要求。目前相关的优化算法有启发式算法和预处理算法两种。启发式算法(a*)［2］使用合适的启发函数减少搜索空间以获得较高的查询效率，启发函数会直接影响最后的计算结果；预处理算法使用点或边标记法、快捷路径法、区块分割法等对网络中的边进行合并和标记以迅速求出最短路径，但需要大量的辅助存储空间。 根据实际交通网络的特点：在主干道上通过的最短路径最多，存在重要的边和点；对于长距离的路径规划，出发点和目标点的中间节点有可能成为最短路径上的点。本文以a*算法为基础，将地图中关键的节点选为地标，并将地标作为启发函数的启发参数来求得路径规划的合理解。为提高长距离路径规划的查询效率，使用分段处理的思想将查询分割为若干子查询，并给出相关的优化方法。 1 相关研究 对于静态网络图g=(v,e)，大多数优化算法需要经过充分的预处理。三角启发算法（a* landmarks triangle，atl）［3］39算法将图g按中心点划分为若干区域，每个区域选取一个标志点(landmark)，根据三角不等式使搜索路