径向基神经网络在高程异常的应用研究
时间:2021-08-15 10:32 来源:毕业论文 作者:毕业论文 点击:次
摘 要GPS 高程测量是利用 GPS 测量技术直接测定地面点的大地高,其较于传统 的水准测量要便捷省时且精度高,人们期望 GPS 测量技术能够逐渐代替传统水 准测量。采用 GPS 测高技术首先要研究 GPS 大地高到正常高的转换,从而对高 程异常问题进行研究。70831 本文在阐述 GPS 高程测量的原理和高程系统理论的基础上,研究分析了高 程转换模型,深入探讨了二次曲面拟合的原理与高程拟合过程。同时为了解决在 高程转换中几何方法的不足之处,本文着重研究了基于人工神经网络的高程拟合 方法。神经网络也称人工神经网络,其本质是以人脑或生物的基本特性为基础建 立数据模型,并对类似于大脑神经突触联接的结构进行信息处理。通过具体的实 例分析,研究径向基神经网络高程拟合过程以及精度分析。同时,将二次曲面拟 合与径向基神经网络进行精度对比,分析两者拟合情况。 该论文有图 11 幅,表 5 个,参考文献 30 篇。 毕业论文关键词:GPS 高程 测量高程异常 二次曲面拟合 径向基神经网络 Application of Radial Basis Function Neural Network in Elevation Anomaly Abstract GPS height measurement is the use of GPS measurement technology directly determine the earth ground point is high, it is convenient and time-saving than traditional leveling accuracy is high, so the GPS measurement technology are expected to gradually replace the traditional leveling. Using GPS height measurement technology must first study the GPS geodetic to normal high conversion, so as to study the problems of the height anomaly. In this paper, on the principle of GPS height measurement and elevation, on the basis of system theory, studied and analyzed the height conversion model, further discusses the principles of quadric surface fitting and elevation fitting process. At the same time in order to solve the geometric method in elevation transform deficiency, this paper focuses on the elevation fitting method based on artificial neural network. Neural network based on the human brain or biological neural network basic characteristics of abstraction and modeling, study the program, adaptability, the essence of the style of brain information processing. Through the concrete instance analysis, study the process elevation fitting RBF neural network and precision analysis. At the same time, the quadric surface fitting accuracy comparing with RBF neural network, the fitting situation are analyzed. The paper has 11 figures,5 tables and 30 references. Key words: GPS height measurement Hight anomaly quadric surface fitting RBF neural network 目 录 摘 要 I Abstract II 目 录 III 图清单 IV 表清单 IV 1 绪论 1 1.1 研究背景意义 1 1.3 文章结构安排 4 1.4 文章技术路线 6 2 高程系统理论 7 2.1 高程基准面 7 2.2 国家高程系统 8 3 高程转换模型 (责任编辑:qin) |