动态环境下CGF实时路径重新规划算法  

Algorithm for CGF Real-time Route Replanning in Dynamic Environment

作　　者：孙少斌[1] 王宽全[2] 林学华[1] 韩志军[1] 

机构地区：[1]蚌埠坦克学院仿真训练中心,安徽蚌埠233013 [2]哈尔滨工业大学计算机科学与技术学院,哈尔滨150001

出　　处：《系统仿真学报》2007年第13期2895-2898,2902,共5页Journal of System Simulation

摘　　要：路径规划是CGF行为模拟最主要和最常用的规划,CGF沿着基于初始信息规划的路径机动时经常会发现路径耗费发生了变化,剩余的路径需要重新规划。D*(动态A*)算法是一个适合于动态环境的实时路径重新规划算法,它通过增量式传播路径耗费的变化提高路径重新规划的效率。介绍了D*算法的一种扩展方法,通过利用问题领域的启发信息引导算法的状态扩展聚焦于当前的状态,减少了状态扩展的数量,进一步提高了CGF在动态环境下的路经重新规划效率。Optimal route planning is central to CGF＇s behaviors simulation. The arc costs may change during CGF traversing through the planed route based on initial information, then the remainder of the path needs to be replanned. The D＊ algorithm （Dynamic A＊） plan optimal route in real-time by incrementally propagating path cost changes as new information is discovered, which is suitable for route replanning in dynamic environment. An extension to D＊ algorithm was described that uses domain heuristic information to guide state expansion, thus reduces the number of state expansion. This extension significantly improves the efficiency of CGF＇s route re-planning in a dynamic environment.

关 键 词：CGF 路径规划 D＊算法 状态扩展 启发信息 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]