首页 | 本学科首页   官方微博 | 高级检索  
     检索      

BP神经网络拟合区域似大地水准面的应用分析
引用本文:聂建亮,郭春喜,程传录,蒋光伟,王斌,马新莹,王文利.BP神经网络拟合区域似大地水准面的应用分析[J].测绘工程,2012,21(1):21-24.
作者姓名:聂建亮  郭春喜  程传录  蒋光伟  王斌  马新莹  王文利
作者单位:国家测绘地理信息局大地测量数据处理中心,陕西西安,710054
基金项目:863计划资助项目(2009AA121402-5);陕测科(2009-03)
摘    要:针对区域重力似大地水准面与GPS/水准数据可能受系统误差影响而降低计算精度的问题,采用BP神经网络拟合区域似大地水准面,削弱系统误差等因素影响,提高区域似大地水准面的拟合效果。分析BP神经网络进行区域似大地水准面拟合的局限性。使用不同区域范围的数据进行验证,结果表明BP神经网络在控制区域内能够较好地逼近实际似大地水准面,而在区域之外具有较差的外推能力。

关 键 词:似大地水准面  BP神经网络  系统误差

The analysis of BP neural network fitting the quasigeoid
NIE Jian-liang , GUO Chun-xi , CHENG Chuan-lu , JIANG Guang-wei , WANG Bin , MA Xin-ying , WANG Wen-li.The analysis of BP neural network fitting the quasigeoid[J].Engineering of Surveying and Mapping,2012,21(1):21-24.
Authors:NIE Jian-liang  GUO Chun-xi  CHENG Chuan-lu  JIANG Guang-wei  WANG Bin  MA Xin-ying  WANG Wen-li
Institution:(Geodetic Data Processing Center,National Administration of Surveying,Mapping and Geoinformation,Xi'an 710054,China)
Abstract:BP neural network is used to weaken the systematic errors in order to improve the accuracy of quasigeoid for the systematic errors in the data of GPS and level and the data of gravity quasigeoid may reduce the reliability.The deficiency for fitting the quasigeoid with BP neural network is discussed.Two sets of data in different areas are employed to test BP neural network.It is shown that BP neural network can factually approach the true quasigeoid,but have poor generalization ability out of control area.
Keywords:quasigeoid  BP neural network  systematic errors
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号