转换GPS高程的神经元网络方法  

Neural Network Method for GPS Height Transformation

作　　者：胡伍生[1] 华锡生[2] 鲍兴南 

机构地区：[1]东南大学交通学院,江苏南京210018 [2]河海大学土木工程学院,江苏南京210098 [3]江苏省地质测绘研究院,江苏南京210008

出　　处：《河海大学学报（自然科学版）》2001年第6期87-89,共3页Journal of Hohai University(Natural Sciences)

摘　　要：介绍了神经元网络BP算法的基本结构 ,提出了转换GPS高程的神经元网络模型———经改进的 5层BP网络结构 .该BP网络结构包括 :输入转换层、输入层、隐含层、输出层和输出转换层 .由于BP网络激活函数的数据限定区间为 [0 ,1],在工程应用中 ,增加设置输入转换层和输出转换层是必要的 .通过某工程实例 ,对 5层BP网络的具体模型结构进行了一些试验研究 ,如输入输出层的结构设计、隐含层最佳节点数的选取等 .最后得到了一些有工程实用价值的结论 .The basic structure of BP algorithm of neural network is briefly introduced, and a neural network model-an advanced five-layer structure of BP network is proposed for GPS height transformation. The present neural network model consists of an input transformation layer, an input layer, an implicit layer, an output layer and an output transformation layer. For ordinary engineering application, the input transformation layer and output transformation layer are needed because the input and output of the Sigmoid standard active function f(x) are limited in the range of 0 to 1. Combined with a practical project, a series of experimental studies are made on the structure of the specific model, such as the structural design of the input and output layer, and the selection of the optimal node number for the implicit layer. Finally some conclusions of practical value are drawn.

关 键 词：GPS高程 神经元网络方法 BP算法 BP网络结构 

分 类 号：P228.4[天文地球—大地测量学与测量工程] TP183[天文地球—测绘科学与技术]