Research On Algorithms Of Earth Rotation Parameters Combine And High Precision Prediction

The earth's movement will be caused by the material flow in all spheres of the earth,and the overall manifestation of the movement is the rotation of the earth which generally described by the earth orientation parameters(EOP).In addition,EOP is also an important parameter to achieve the mutual transformation between the terrestrial and celestial sphere reference frame.High-precision earth orientation parameters are essential basic data for the precise orientation and positioning of the earth's reference system and ground observation stations in the space of the universe,satellite precision orbit determination,autonomous navigation of high-altitude detectors,and high-precision timing.Currently,EOP is often provided by the International Earth Rotation and Reference Systems Service(IERS),which is a combination of VLBI,GNSS,SLR,and DORIS observations.Due to the different observation methods and principles the solved parameters were bound to has difference.Therefore,it is necessary to consider the problem of systematic differences between different technologies when combine the results of various technologies.A reasonable and combine method is the focus of research for obtaining the final results.With the application of space measurement technology in ERP measurement,the accuracy of ERP observation is significantly improved,but the real-time requirements of space location technology can not be satisfied in the solution cycle.Therefore,this problem can only be solved by the prediction method.In this paper,the above problems are studied on the combine of earth rotation parameters and high-precision prediction theory and algorithms.The main contents and results of the paper include:1.For the problems of the ERP results published by IERS are not consistent with the epochs published by each technical data analysis center,this paper uses the results of the earth's rotation parameters published by GNSS technology integrated data processing center(IGS),using the two methods of parameter transformation and parameter adjustment with restriction conditions,the single technical results are estimated to the same epoch of IERS,and the results are compared with the results of IERS C04.The calculation results show that the results from the two approaches are the same,but the parameter conversion method is more simple than the parameter adjustment method with restrictions.2.The residual series obtained by the above two methods is not a stable,zero-mean sequence,and there are still significant high-frequency changes.The spectrum analysis is used to obtain the periodicity in the residual series,and the least squares fitting is used to eliminate the period term in the residual series.Compared with the original residual series,the residual accuracy has a certain improvement.3.In view of the different data processing software and processing methods used by different GNSS analysis centers,the final combined solution may have system error.Therefore,the results of each analysis center were weighted by their variances and combined model with additional benchmark constraints,the results of ERP parameters were combined compared with the IERS C04 results.The results show that the accuracy of the results calculated by the additional boundary constraint fusion model is significantly higher than that of the weighted average results.The former results are equivalent to the accuracy of the IGS fusion results.4.Taking into consideration the different accuracy of polar movement observations in different time periods,using robust estimation methods to correct the part of the least squares fitting term,a prediction model of RLS+AR is proposed.The results show that the prediction accuracy of the RLS+AR model considering the difference in accuracy before and after the basic sequence is significantly higher than that of the traditional LS+AR model.In order to improve the efficiency of prediction,the LS+AR prediction model is used to predict the polar movement parameters according to the basic sequence of different lengths.The results show that the accuracy of the polar movement forecast with a base sequence length of 10 years is the highest,and the longer the basic sequence is,the worse the forecast results are.