Research on the GPS Elevation Fitting with RBF Neural Network Model Considering Effects of Sample Data Preprocessing

Abstract:

RBF neural network and three kinds of preprocessing methods are introduced, and this paper used these preprocessing methods combined with RBF neural network and strict RBF neural network to perform elevation fitting. Comparing and analyzing the fitting results, the results show that preprocessing methods can affect elevation fitting results. Centralized preprocessing data maximum improves RBF neural network elevation fitting precision, and it also let RBF neural network have stronger generalization ability. Normalization preprocessing methods are not necessarily optimal. It is essential for us to choose preprocessing method to fit the elevation.

• Periodical: Applied Mechanics and Materials (Volumes 568-570)
• Pages: 817-821
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.568-570.817
• Online since: June 2014
• Authors: Li Long Liu, Jun Yu Li*, Chen Hui Cai, Guo Biao Lin
• Keywords: Centralization, Elevation Fitting, Modeling Training, Preprocessing, RBF Neural Network

References

[1] Liu Jian, Wang Qijie, Wang Xiaohui, et al . Application of RBF neural network to GPS elevation fitting based on genetic algorithm [J] . Science of Surveying and Mapping, 2013, (2): 143-145. (in Chinese).

[2] Ren Chao, Wu Wei, Huang Zhengkai, et al . RBF neural network application in GPS height fitting based on AIC criterion [J]. Science of Surveying and Mapping, 2013, (2) : 77-79. (in Chinese).

[3] Zhang Hao, Wang Qijie, Zhu Jianjun, et al . INFLUENCE OF SAMPLE DATA PREPROCESSING ON BP NEURAL NETWORK-BASED GPS ELEVATION FITTING [J] . JOURNAL OF GEODESY AND GEODYNAMICS, 2011, (2): 125 - 128. (in Chinese).

[4] Wang Xiaohui, Wang Qijie, Ding Yuanlan et al. COMBINED MODEL IN HEIGHT ANOMALY FITTING, [J]. JOURNAL OF GEODESY AND GEODYNAMICS , 2012, (6): 103-106. (in Chinese).

[5] Chen Ming. MATLAB neural network theory and examples of fine solution [M] Beijing: Tsinghua University Press, 2013. (in Chinese).

[6] Ji Zhangjian, Yuan Yunbin, Sheng Chuanzhen. GPS HEIGHT FITTTING BASED ON CHAOS PRATICLE SWARM SUPPORT VECTOR MACHINE AND CONSIDERING EFFECT OF TERRAIN [J]. JOURNAL OF GEODESY AND GEODYNAMICS, 2010, (2): 95 – 98. (in Chinese).

[7] Huang Liangke, Liu Lilong, Tang Yanxin et al. Area Quasi-Geoid Refine Based on Improved Kriging Method[J]. Journal of Guilin University of Technology, 2012, (4) : 532-536. (in Chinese).

[8] Sun Chuansheng, YANG Guodong, Wu Qiong. The Application of ArtificialNeural Network to GPS Height Elevation fitting [J]. Bulletion of surveying and Mapping, 2011, (8) : 48-50. (in Chinese).