目前導航定位普遍使用的技術為全球定位系統(GPS)，由於使用在動態環境下，且必須考量到設備成本問題；因此，導航使用GPS單點定位並以電碼觀測量求解坐標。受到定位誤差的影響，一般單點動態定位精度約15至25公尺，而考量到未來車用導航應用如自走車、行車安全系統及空間資訊服務等發展趨勢，這些項目至少要公尺等級的定位精度才能達到。

因此，提供給導航使用者即時差分改正量，將有助於定位精度的提升，也能順應未來導航發展趨勢。美國WAAS可提供北美地區即時電碼的差分改正，定位精度可達1至5公尺，但只侷限於北美洲，台灣目前並無星基增強系統，而考量到基地站及接收儀設置成本，類似WAAS電碼差分定位不符合區域型導航的用途。本研究提出利用類神經網路發展區域型即時動態單點定位的差分改正演算法，由於差分改正量為時間與空間分布之相關的函數，因此，希望以人工智慧的方式去學習，讓使用者能即時獲得差分改正量，進而提升定位精度。

由實驗結果可知，本研究所提出使用類神經網路預估差分改正量以提供區域型的差分改正模式，確實可提升單頻電碼單點定位精度。對於導航使用者來說，在不需要架設主站前提下可獲得即時差分改正量得到高精度定位成果，且不需要額外的硬體設備，解決以往由無線通訊設備傳送改正量遭遇的訊號中斷或遮蔽問題，可有效降低成本花費，較符合導航使用者的需求，對於未來導航的發展也有幫助。

Globe Positioning System (GPS) is the primary navigation technique used in land vehicular navigation applications nowadays. Complicated environments during navigation and the cost of equipment become a big issue, therefore, the single point positioning mode and code measurements are applied for kinematic positioning. The land vehicular navigation system plays an important role in the development of a modern vehicle, for example: the auto-drive vehicle , the system of safety driving and geometry information service and so on. If we want to achieve these purposes, the accuracy of single-point positioning must be in meter level, which will be affected by the various error sources of GPS.

Therefore, if the real-time differential correction can be applied to vehicular navigation users, the positioning accuracy can be improved and to catch the trend of land vehicular navigation applications Today, USA’s WAAS offers real-time differential correction to unlimited number of users only in the vicinity of North America and improve the positioning accuracy from one to five meters but only in North America. However, there are no such real-time differential correction services available in Taiwan; besides, WAAS is not suitable for users in the vicinity of Taiwan as its differential corrections are not valid in this region. . In the thesis, the development of artificial neural networks based GPS differential correction algorithms for single point positioning is implemented. The proposed ANN based algorithm is applied to generate real time differential correction to improve the positioning accuracy of the test sites because those differential corrections can be considered as the function of time and spatial distribution, therefore, we can use.

The preliminarily results presented in this study indicate that the utilization of ANN based regional differential correction model does improve the accuracy of conventional signal point positioning. For vehicular navigation users, they don’t need to receive external information from reference station thus the extra cost of equipment can be eliminated. In addition, the proposed algorithm can get the same level of accuracy as the use of conventional code differential positioning as well as solve the problem concerning the outage or jamming of broadcasted correction signal when using the wireless communication equipment to broadcast the differential correction to rover station. Therefore, the findings of this research can be considered significant concerning the accuracy improved and potential saving of hardware expanse for developing future land vehicular navigation systems.

1.內政部國土測繪中心e-GPS即時動態定位系統入口網站。http://www.egps.nlsc.gov.tw/index.html. (摘於2009年5月8日)。
2.李征航(1996)。高新技術講座，全球定位系統技術新進展：第二講差分GPS。武測科技，第一期，1996年，41-48。
3.李坤煌(1996)。GPS軌道誤差特性及基線重複性分析。國立交通大學土木研究所碩士論文。
4.曾清涼、余致義、何慶雄、劉啟清、楊名(1997)。各類儀器之操作說明。"GPS衛星定位測量實務"(曾清涼、余致義、何慶雄、劉啟清、楊名編著)。成功大學衛星資訊研究中心，台南。
5.曾清涼、儲慶美(1999)。GPS衛星測量原理及應用。成功大學衛星資訊研究中心，台南。
6.楊名、江凱偉(2008)。全球導航衛星系統(GNSS)資料聯合處理技術。內政部國土測繪中心委託計劃期末報告修正本。國立成功大學測量及空間資訊學系，台南。
7.劉基余、李征航、王躍虎、桑吉章(1993)。全球定位系統原理與其應用。測繪出版社，北京。
8.儲慶美(1995)。全球定位系統之原理與應用。第23屆國軍軍事著作金像獎作，11月。
9.100gogo. http://www.100gogo.com/ever1.htm. (Accessed on 3, May, 2009).
10.Abidin, H.Z. (1992). Some aspects of on-the-fly ambiguity resolution. Proceedings of the Sixth International Geodetic Symposium on Satellite Positioning, 660-669.
11.Anderson, D. and McNeill, R. (1992). Artificial Neural Network Technology, Technical Reports, ITT Industries.
12.Bagley, L.C. and Lamons, J.W. (1992). NAVSTAR joint program office and a status report on the GPS program. Procs. 6th Int. Geodetic Symp. on Satellite Positioning, Columbus, Ohio, 17-20 March ,1992 ,21-30.
13.Bishop, C.M. (1995). Neural Networks for pattern Recognition, Oxford University Press.
14.Cawsey, A. (1998). The Essence of Artificial Intelligence, Prentice Hall PTR.
15.Chiang, K.W. (2004). INS/GPS Integration Using Neural Networks for Land Vehicular Navigation Applications. Department of Geomatics Engineering, The University of Calgary, Calgary, Canada, UCGE Report 20209, Status.
16.Chu, C.M. (1993). Efficient and Effective Handling of Cycle Slips in Global Positioning System Data. PhD thesis, School of Surveying, the University of New South Wales, Sydney, Australia. December.
17.DANA, P.H. (1999). The University of texas, Internet.
18.GIS Development: The Geospatial Resource Portal. http://www.gisdevelopment.net/technology/gps/ma0782.htm. (Accessed on 5, May, 2009).
19.Green, G..B., Massatt, P.D. and Rhodus, N.W. (1989). The GPS21 primary satellite constellation. Navigation, Vol.36, No.1, Spring, 9-24.
20.Ham, F.M. and Kostanic, I. (2003). Principles of Neurocomputing for Science and Engineering, McGraw-Hill.
21.Haykin, S. (1999). Neural Networks: A Comprehensive Foundation, 2nd Edition.
22.Hopfield, J.J (1982). Neural networks and physical systems with emergent collective computational abilities. Proceeding of the National Academy of Science, 2554-2558.
23.Hopfield, H.S. (1970). Tropospheric Effect on Electromagnetically Measured Range: Prediction from Surface Weather Data. Applied Physics Laboratory, Johns Hopkins University, Baltimore, MD.
24.Indriyatmoko, A., Kang, T., Lee, Y.J., Cho, Y.B., Jee, G.I., Kim, J. (2008). Artificial neural networks for predicting DGPS carrier phase and pseudorange correction. GPS Solution.
25.Kayton, M. and Fried, W.R. (1996). Avionics Navigation Systems. John Wiley & Sons, Inc.
26.Master, T. (1995). Neural, Novel & Hybrid Algorithms for Time Series Prediction, John Wiley & Sons, Toronto.
27.Minsky, M. and Papert, S. (1969). Perceptrons, MIT Press.
28.MITRE CAASD. http://www.mitrecaasd.org/proj/satnav. (Accessed on 5, May, 2009).
29.Parkinson, B.W., Spilker, J.J., Axedlrad, P., and Enge, P. (1996). Global Positioning System: Theory and Application Vol I, II. American Institute of Aeronautics and Astronautics.
30.Pfost, D., Casady, W. and Shannon, K. (1998). Precision Agriculture: Global Positioning System (GPS). University of Missouri, Columbia.
31.Rao, B.R.K., Sarma, A.D. and Kumar, Y.R. (2006). Technique to reduce multipath GPS signals. Current Science, Vol. 90, No. 2, 25 January.
32.Rizos, C. (1996). Priciples and Practice of GPS Surveying. School of Geomatic Engineering, The University of New South Wales, Sydney NSW, Australia.
33.Rizos, C. and Grant, D.B. (1990). Time and the Global Positioning System. In “Contributions to GPS Studies”(ed:Rizos, C.), UNISURV S-38, the School of Surveying, University of U.S.W., Australia, 45-101.
34.Rumelhart, D.E., Hinton, G.E. and Williams, R.J. (1986). Learning representations by back-propagation errors. Nature, 533-536.
35.Sasstamoinen, J. (1972). Atmospheric Correction for the Troposphere and Stratosphere in Radio Ranging of Satellites, Geophysical Monograph 15, American Geophysical Union, Washington DC.
36.Seeber, G.. (1993). Satellite Geodesy: Foundations, Methods, and Applications. Walter de Gruyter & Co., Berlin, Germany.
37.Wikipedia. http://en.wikipedia.org/wiki/Artificial_neural_network. (Accessed on 20, May, 2009).
38.Wisconsin State Cartographer’s Office. http://www.sco.wisc.edu/gps/system.php. (Accessed on 3, May, 2009).
39.Wells, D., Beck, N., Delikaraoglou, D., Kleusberg, A., Krakiwsky, E.J., Lachapelle, G.., Lamgley, R.B., Nakiboglu, M., Schwar, K.P., Tranquilla, J.M., and Vanicek, P. (1987). Guide to GPS Positioning, Canadian GPS Associates, Federicton, New Brunswick, Canada.