基于多重分形的地形辅助导航路径规划算法  

Flight route planning for terrain navigation using multi-fractal theory

作　　者：孙立国[1] 李世丹[1] 王德生[1] 赵晓林[1] 李欣[1] 

机构地区：[1]清华大学电子工程系,北京100084

出　　处：《清华大学学报（自然科学版）》2011年第1期111-114,121,共5页Journal of Tsinghua University(Science and Technology)

摘　　要：地形导航是一种利用航行器下方地形信息获取航行器位置的导航技术。为了得到较高的定位精度并使飞行长度尽可能短,需要预先规划飞行路径。该文将地形数字高程图(digital elevation model,DEM)看作二维灰度图像,引入在图像处理中得到广泛应用的多重分形理论,将利用小波变换求得的奇异指数作为地形提供定位精度能力的一种度量。选择合适的阈值划分得到定位优区,利用A*算法寻找最短路径,此路径即是规划路径。选用粒子滤波作为导航算法,沿规划路径和其他路径进行多次飞行仿真,结果显示规划路径的定位精度和稳定度都比较高,而且路径长度短,有较强的实用性。Terrain navigation systems obtain navigation information from the terrain to correlate the aircraft flight path.Accurate positioning and shorter routes then require a flight route plan.A flight planning system was developed using digital elevation model（DEM） analyzed using multi-fractal theory to measure the amount of terrain information with using a singularity exponent as the criterion.The optimal flight route can be determined within suitable areas with the A＊ algorithm.This analysis used a particle filter for the terrain navigation algorithm.Comparison of the planned flight route and other flight routes in numerical simulations showed that the planned flight route was indeed the shortest,with the results showing the system was accurate and stable.

关 键 词：图像处理 多重分形 小波变换 奇异指数 A＊算法 粒子滤波 

分 类 号：TP751.1[自动化与计算机技术—检测技术与自动化装置]