GPS HEIGHT CONVERSION BASED ON COMBINATION METHODSOF SURFACE FITTING AND BP NEURAL NETWORK
Hou Dongyang 1) ; and Zhang Shubi 1,2)
1School of Environment Science and Spatial Informatics of CUMT, Xuzhou 2210082Key Laboratory of Resources and Environmental Information Engineering of Jiangsu Province, Xuzhou 221008
Abstract：In the GPS height conversion, it is unable to obtain high accuracy only using surface fitting method and there exist great risk by using BP neural network. To solve the problems, six kinds of combination methods based on the surface fitting and neural network are designed.These methods are compared and analyzed through one example. The results show that the accuracy of the six combination methods are higher than by use of single surface fitting method or BP neural network method, especially the threeinput surface fitting method has highest accuracy.
