GPS ile nirengi ağı sıklaştırmasında uygun yönteminin araştırılması

Abstract

Bu çalışmada GPS ile nirengi ağlarının sıklaştırılması incelenmiş ve günümüz koşullarında Türkiye'de uygulanabilecek ölçme ve değerlendirme yöntemleri için öneriler getirilmiştir. GPS ile ilgili bilgiler genel olarak verilmiş, ölçmelerde sonuçlan etkileyen hata kaynaklan araştırılmış ve bunları gidermek için izlenmesi gereken yol belirtilmiştir. Doğru sonuçlan elde edebilmek için, ölçülerin değerlendirilmesinde yapılması gereken işlemler adım adım açıklanmıştır. Farklı yazılımların değerlendirme yetenekleri incelenmiş, bu yazılımlarla elde edilen sonuçlar karşılaştırılmıştır. Uydu yayın yörünge bilgileri ve prezisyonlu yörünge bilgileri ile yapılan değerlendirmelerin sonuçlara etkisi incelenmiş, sıklaştırma çalışmalarında yayın yörünge bilgilerinin yeterli olacağı görülmüştür. GPS baz çözümü sonuçlarının dengelenmesi ve sonuçların ne şekilde test edileceği belirtilmiştir. Elde edilen GPS sonuçlarının yersel sisteme dönüşümünde, eşlenik noktaların seçimi ve dönüşüm işlemindeki sorunlar incelenmiş ve distorsiyonlu ağlarda uygulama esasları açıklanmıştır. Ayrıca GPS ile oluşturulması gereken Ülke Temel Ağı için değerlendirme ve uygulama esasları önerilmiştir.

The main objective in geodesy is to construct a nationwide network and to densify the network for various purposes. Traditionally, the objective mentioned above is accomplished with the measurement of angles and distances by classical instruments. it is obvious that the classical technique depends on inter visibility between the points and atmospheric conditions. Moreover, measuring with the classical technique may not be convenient for night time. in order to accomplish the same objective, öne can also use the Global Positioning System (GPS), which is introduced at the end of the 1960's. At this point, we note that GPS get rid of most of the disadvantages of the classical techniques. That is, GPS measurements are independent of ali whether conditions, visibility condition betv/een the points is not required, and day-and-night measurements are possible. Because of the advances in the satellite manufacturing technologies and new established satellite systems, GPS has became so popular since 1980. Other important reasons for GPS being common are the following:. GPS is quite proper in determining the precise positioning of a geodetic points.. GPS is an extremely effective method for densification of the geodetic network.. GPS measurements are not affected from the personal errors in contrast to the classical measurements. xi . GPS is the most economical and the fastest way of positioning geodetic points. Therefore, GPS strongly supports network densification. To summarize, GPS supports ali three dimensional positioning. Classical geodetic networks are the main reference systems in national applications. GPS measurement technique supports these systems without using classical techniques. At this point, it is very important to note that these two methods have their own coordinate systems. in order to obtain national system coordinates with GPS, the transformations of the coordinate systems berween these two (Classical technique and GPS) are unavoidable. There are some majör geodetic problems in istanbul that many local reference coordinate systems are in use. Whereas, there should be öne unique reference coordinate system. in addition, the transformations among these coordinate systems have not yet developed fully. Apart from these, each local reference coordinate system has a large distortion even in close regions. in 1987, a sophisticated geodetic work has been done for establishing a precise geodetic network, istanbul Metropolitan Netv/ork (IMN) in national datum (the average positioning rms. is less then 3cm) The work mentioned above was directed by A. Aksoy. in this work, firstly, we computed the transformation parameters with the minimum set of GPS measurements for different models. Secondly, we tested them in the difFerent part of the network. Thirdly, we discussed the best way of accomplishing this objective. For detail analysis of the GPS methods, we use IMN. IMN based on the 12 known points and 159 new points in the national datum. The maximum point positioning rms of the national datum is less then 3 cm. On the other hand, there is no precise height xii information. Here, the height computed from zenith angles measurements. Fortunately, a few points of the network are constructed by spirit leveling based on national vertical datum. Also, except two points, there is no precise knowledge about geoid height in the national datum. The name of these points are 34056 and 34071. When we tried to obtain GPS measurements at these points, the measurements was not successfiılly performed at the point 34071 due to the bad sky visibility and telecommunication antennas, which were surrounding the point. We made measurements at 6 points of the network as a closed traverse by means of rvvo Ashtech Z XII receivers. These receivers can receive L l and L2 frequencies transmitted from the satellites. in 1995, we measured 9 baseline from 051 GPS day 055 GPS day. The length of the shortest baseline is about 11 km and the longest öne is a little more than 30 km. The measurements at each point, lasted from l hour to 2 hours, were performed in day time while at least 4 satellites were visible. Except the points 34056, 34052, and 34013, the antennas mounted on the pillars. At these points a tribrach was used and the antenna height was measured precisely(about l mm). Atmospheric measurements of atmospheric pressure, dry and wet temperature were performed at the beginning and at the end of the each session. in order to test the transformation results, which are öne of the majör parts of this study, 4 additional network points are included and l hours GPS measurements were performed at these 4 points. These points are numbered as 34220, 34218, 34121 and 34112. in addition, a kinematics measurements campaign has been performed for the test netvvork, located in ITU campus. The procedure of static initialization, antenna swapping and rapid static were performed as initial measurements. The kinematics data obtained from these three ways is processed with Prism softvvare. xiii Post processing was performed with two different software; Bernesee version 3.5 and Ashtech Prism. The results are shown in Section 6 explicitly. From these results, we notice that there is no difference in baseline component results using the atmospheric measurements in the standard atmosphere condition for troposphere corrections. Although the LI solutions with two software are the same, L3 solutions with these two software resulted some small differences. The L3 solutions is performed as campaign solution using broadcast ephemerides and IGS precise ephemerides. The result, given in section 6, shows that there is no difference using precise and broadcast ephemerides. There is only a few mm difference and a little good rms when the precise ephemerides solution results are used. The LI solution obtained with ionosphere model correction is compared to L3 solution. It is found that there is a difference of maximum 9 mm in the Cartesian coordinates. Moreover, the LI solution obtained without ionosphere model correction is compared to L3 solution and it is seen that there is a difference of 2.4 cm in the Cartesian coordinates. The L3 solutions with precise ephemerides were used for other applications as datum transformation. The datum transformation is achieved successfully in two and three dimensions. The following transformation methods were used between the GPS system and the local geodetic system. A similarity transformation is used in two dimension and the following three dimensional methods are examined and compared each other:. Bursa- Wolf method. Moledenski-Bedekas method. 1 0 parameter Afifine method. Veis method The result obtained by using two and three dimensional transformation methods are almost the same. xiv As is known, for the densification process, one must sure that the known points in both systems (namely, GPS and local system) have to enclose the new constructed points. If this is the case, the transformation works quite well. Otherwise, which is the case when the new points are outside the network, the transformation parameters, obtained earlier, should be applied to the GPS vectors instead of GPS coordinates. Here, the GPS vector are defined as the vector from the known point to the measured new point. The transformation process which is defined above, is completed by adding transformed vectors to the related known local coordinate. In addition, one should use a test technique for detecting the outliers. For this purpose, we examined and compared the existing outlier detection methods and we observed that detecting the outliers in transformation is quite difficult problem. Because a coordinate has two or three components (x, y) or (x, y, z) respectively and if only one component of a coordinate is detected as an outlier, what should be the strategy whether this point is included or excluded from the solution. Benning introduced a test method based on the Kock's idea for two dimensional coordinate transformation. We extended and developed this idea to the test method for three dimensional coordinate transformation. We thus obtained a straight forward strategy of detecting the ouliers for three dimensional coordinate transformation.