GPS高程拟合方法研究

Research on GPS Elevation Fitting Method

GPS测量获得的高程数据是以WGS-84椭球为基准的大地高,而工程中采用的是以似大地水准面为基准的正常高,两者之间存在高程异常。当前,主要采用GPS高程拟合求取高程异常值,而利用GPS高程拟合获得满足一定精度的正常高是测绘行业的研究热点之一。本文通过某测区GPS控制网数据与水准测量数据,采用多项式曲线拟合法、多面函数拟合法、二次曲面拟合法、RBF神经网络拟合法、二次曲面拟合法与RBF神经网络组合拟合法求取高程异常值。结果表明:1)单一拟合法的拟合精度低于组合拟合法拟合精度。2)依据内、外符合精度,本文采用的5种高程拟合方法均满足四等水准测量要求。因此,在精度允许的前提下,可借助GPS高程拟合方法代替传统水准测量,进行高程值的获取。研究成果对GPS高程拟合在实际工程中的应用具有重要参考价值。

The elevation data obtained by GPS measurement is the ellipsoidal height based on the WGS84 ellipsoid,while the engineering uses the orthometric height based on quasi-geoid.There is an elevation anomaly between them.At present,GPS elevation fitting is mainly used to obtain elevation anomaly values,and the use of GPS elevation fitting to obtain a orthometric height that meets certain accuracy is one of the research hotspots in the surveying and mapping industry.In this paper,we use the polynomial curve fitting method,the multi-faceted function fitting method,the conicoid fitting method,the RBF neural network method,and the conicoid and the RBF neural network combination fitting method to obtain the elevation anomaly value through the GPS data and leveling data of a surveying area.The results show that:(1)The fitting accuracy of single fitting method is lower than that of combination fitting method.(2)According to the internal and external conformity accuracy,the five elevation fitting methods adopted in this paper all meet the accuracy requirements of fourth-order leveling.Therefore,under the premise of acceptable accuracy,the GPS elevation fitting method can substitute the traditional leveling to obtain the elevation value.The results have important reference value for the application of GPS elevation fitting in practical engineering.