不同植被GPS定位精度影响因素分析及多路径效应

Influencing factors analysis and multi-path effects of different plant cover types on GPS positioning accuracy for forest fire control

针对不同植被类型森林火灾的灾前预警、灾中扑救、灾后评估过程中需要精准确定火灾、水源位置和扑火路线的需求，利用Trimble AgGPS132和Geo XT进行了草地（鸭趾草）、灌木丛（大叶黄杨）和树林（落羽杉）等不同植被类型时的GPS静态定位和动态定位试验，并观测记录了GPS卫星可见性和PDOP，对比分析其定位误差，进行不同截止高度角对能见卫星个数、PDOP和定位误差的影响，并分析其误差原因及改善方法，采用AUTOCAD结合速度过滤算法对不同植被类型时动态定位产生的多线进行简化。试验结果显示，根据卫星可见数和PDOP值，不同植被类型的GPS观测条件为草地〉灌木丛〉树林，定位误差的大小为草地〈灌木丛〈树林，灌木丛、树叶产生的多路径效应是在森林中定位误差的主要来源。得到了不同植被类型下测量次数与定位精度的回归拟合方程，定位精度与测量次数的对数成线性关系，静态定位时测量次数为200次比较合适。截止高度角太大，将导致观测失效、误差增大的现象，综合考虑根据卫星可见数、PDOP值、定位误差和减弱森林中的多路径效应等因素，截止高度角E在8°～20°比较合适。另外对于草地、灌木丛、树林不同植被类型进行动态定位时，速度过滤算法都能起到很好的修正误差、优化线型的作用。

In the evaluation processes of pre-disaster warning, firefighting and post-disaster assessment, in order to meet the needs of accurately determining forest fire, water source position and firefighting routes, the experiments of GPS static and dynamic positioning were conducted by using Trimble AgGPS132, to test the plant cover type＇ s effects （three covers were tested： grassland, coppice and deciduous forest） on GPS satellite visibility, PDOP and positioning errors. The stationary test of the influence of recording number on positioning accuracy was conducted for various plant cover types. The field trail was also carried out to compare the effects of different elevation mask angles on number of satellites used, PDOP and positioning error. The mobile test was to examine the accuracy of GPS signals in different site by handheld device-Trimble Geo XT, the velocity filter algorithm was combined with Auto CAD to fulfill the task of polyline simplification. The experiments were also carried out to compare the effect of different elevation mask angles on number of satellites used, PDOP and positioning error. According to the results, both the plant cover types and elevation mask angles affected the satellite visibility, PDOP and positioning error. Based on the satellite visibility, PDOP value, the sequencing of surveying condition under different plant cover types was as follows： grassland 〉 coppice 〉 deciduous forest. Moreover, the sequencing of positioning error under different plant cover types was as follows： grassland 〈 coppice 〈 deciduous forest, and this is because the multipath effect of signal caused by the tree stand is the main source of positioning error. For the stationary test, the positioning errors decreased linearly with the logarithm of recording number taken into account, so the appropriate static positioning measurement times was 200 times. It showed that the large setting elevation mask angle lead to failure survey and big error. Finally, the optimal value （8 ° ～ 20° ） of elevation mask angle was implemented. For the mobile test, velocity filter algorithm contributed to produce smother liners and straightening the sequences of positioning points.