Research On The Data Fusing Of Terrestrial Reference Frame At Normal Equation Level And Its Application In Grid Computing

This dissertation mainly focuses on the fusing methods of Terrestrial Reference Frame(TRF) data and its application of Grid Computing at normal equation level, including the theory of adjustment datum, the datum implementation and realization of TRF, and TRF Grid-Computing system, and so on. The main works and contributions are summarized as follows:1. Concludes least squares adjustment models at normal equation level. The adjustment datum is analyzed in algebra. The centrobaric datum space is the orthogonal complement of the rank-defect normal matrix space. Through a leveling network test, the accessibility of adjustment and datum transformation based on normal equation is validated.2. Discusses the similarity transformation model s application used in the datum implementation of TRF,the accuracy of ITRF2005 is evaluated using four sets of station coordinates of space geodetic techniques GPS SLR DORIS VLBI at 2000.0 epoch, by experiment.3. Deduces the Adding and Removing Algorithm of constraints rigorously for providing the reference on SINEX format. Aiming at the basic requirement of maintenance of TRF this paper analyses the deficiency of data-processing for regional GPS Geodetic Network(Geo-Net) currently. Also, we describe the minimum constraints (MC) approach to solve or reconstruct the solution of regional GPS Geo-Net. We discuss the character of the MC approach through comparing several constrained solutions and point out that the MC Approach is an optimal method to express the regional GPS Geo-Net in TRF finally.4. From the Similarity Transforming model, we construct a two-level relation model of combined adjustment. The two key problems, single solution processing and datum definition in TRF construction, are analyzed particularly. Using more than two years week-time-series, we get the Similarity-Transform and Minimal fusing solution of EPN achieve the Europe GPS Reference Frame ,involving 38 geodetic points, in cm level.5. We constitute a pilot TRF Grid-computing system. The Grid-computing platform, Meta algorithm and its fusing algorithm are constructed. The simulating test indicates that the Grid-Computing can complete the compute and release of the TRF products in a perfectly efficient.