Orhan AKYILMAZ, Tevfik AYAN

Esnek hesaplama yöntemlerinin jeodezide uygulamaları

Bu çalışmada Yapay Sinir Ağları (YSA) ve Bulanık Çıkarım Sistemleri (BÇS) jeodezinin en önemli problemlerinden bazılarının çözümü için kullanılmıştır. Bu problemler sırayla, Yer yuvarı gravite alanının modellenmesi ve de GPS/Nivelman ölçülerinden geoid yüzeyinin belirlenmesidir. Gravite alanı modellemesi için GRACE (Gravity Recovery And Climate Experiment) uydu gravimetre verileri.GPS/Nivelman geoidi için ise İzmir metropolitan GPS nirengi ağında yapılan GPS ve nivelman ölçülerinden türetilmiş geoid yükseklikleri bilinen kontrol noktaları kullanılmıştır. Her iki uygulama sonucunda esnek hesaplama yöntemlerinin, özellikle girdi-çıktı sistemleri şeklinde tanımlanan jeodezik problemlerin çözümü için uygun yöntemler olduğu sonucuna varılmıştır

Application of soft computing methods in geodesy

Soft computing methods such as artificial neural networks (ANN) and fuzzy inference systems (FIS) have been widely used methods in various science and engineering fields. As the backgrounds of these methods are not very old, they have shown a rapid development with the improvements in computer systems and computation techniques. Their use in geodesy is quite new. In this study, both ANN and FIS have been used to solve the some of the major problems in geodesy. These problems are modelling of Earth's gravity field and the determination of geoid surface from GPS/Levelling. For gravity field modeling, GRACE (Gravity recovery And Climate Experiment) satellite gravimetry data, and for GPS/Levelling geoid control points with known geoid heights derived from GPS and levelling measurements in İzmir metropolitan GPS network were used The results from the soft methods used in computations were also compared with those from the conventional methods in terms of model quality measures like root mean square (RMS) error, mean error, mean absolute error, error range, correlation coefficient. Both applications' results have concluded that soft computing methods are appropriate methods for the solution of geodetic problems especially which can be defined as input-output systems and should be considered for the solution of different other problems in geodetic science.

PDF

___

Akyılmaz, O., Ayan, T., Özlüdemir, M. T., (2003). Geoid surface approximation by using Adaptive Network based Fuzzy Inference Systems. AVN, 8-9, 22-27, Wichmann, Hüthig, Almanya.
Gruber, T., (2002). Gravity Field Models beyond CHAMP, GRACE and GOCE: A synergetic view of global gravity field computation. Proc. of Heiskanen Symposium in Geodesy, 1-4, Columbus, OH., ABD.
Han, S.-C., (2003). Efficient global gravity determination from satellite-to-satellite tracking (SST), PhD. Thesis, OSU, Columbus, OH., ABD.
Haykin, S.S., (1994). Neural Networks: A Comprehensive Foundation, NY: Mcmillan Press.
Jang, J.-S.R., Sun, C.-T., Mizutani, E., (1997). Neuro- Fuzzy and Soft Computing: A Computational Approach to Learning and Machine Intelligence, First Edition, Prentice Hall.
Moritz, H., (1992). Geodetic reference system 1980. Bull. Géod. 66, 2, 187-192.
Rapp, R.H., (1986). Global geopotential solutions, In: Lecture Notes in Earth Sciences: Mathematical and Numerical Techniques in Physical Geodesy, pp. 365-415, Eds. Sünkel H., Springer-Verlag, Heidelberg, Almanya.
Torge, W., (1975). Geodaesie. Sammlung göschen, walter de gruyter, 268 s., Berlin-New York, ABD.
Zadeh, L.A., (1965). Fuzzy sets. Information and Control, 8, 3, 338-353.