Research On Systematic Errors Processing Method And Application Of Partial EIV Model

With the development of geodesy,data acquisition means constantly enrich and data precision is gradually improved,how to obtain useful data from these data,to obtain reliable adjustment results and meet the actual measurement precision information required are the hot spots in the field of current measurement data processing.However,the methods of systematic errors processing are almost limited to the Gauss-Markov model.In fact,the coefficient matrix is also affected by random errors.Therefore,improving the method of systematic error and enriching the total least squares theory are the major researches of this paper.The detailed researches are summarized as follows:Considering the case of the random errors in the coefficient matrix,the solutions of the Partial EIV model with additional systematic parameters are derived.And the adjustment model of the additional systematic parameters of the scale parameter in the current level measurement is reconstructed to the Partial EIV model of additional systematic parameters on the scale parameter.The results of real example and simulated example show that the effectiveness of this method is basically the same as that of traditional method in dealing with scale parameter as additional systematic parameters,and the related analyses are given.In order to weaken the bad effect on the estimated values,the Partial EIV semi-parametric model is built.The penalized least square criterion is introduced to derive formulae,and choose the proper regularization matrix in the experiments.The examples show that the method is effective and feasible.This paper adds non parametric part(systematic errors)in the Partial EIV model,building the Partial EIV model with systematic errors.And the model is rewritten as a nonlinear model.The formulae of solutions and second order bias and covariance of parameter estimation are derived based on the penalized total least squares criterion.Then the mean square error of parameter estimation is used as the index of precision estimation.In the calculation,the selection of the smoothing factor is determined by the U-curve method.The results of experiments show that the proposed method is superior to the traditional method in terms of parameter estimation results and the mean square error.It validates the algorithm can mitigate systematic errors to a certain extent further and the feasibility and effectiveness of the proposed method.Compared with the L-curve method,the results obtained by the U-curve method are more accurate.The systematic errors processing method of Partial EIV is applied into the multibeam echosounding data processing.Considering that the multibeam echosounding data are mainly influenced by the attitude error and the sound velocity measurement errors,the inevitable systematic errors are produced.And it takes into account that the trend term is fitted by polynomial,the coefficient matrix is composed by observations.Then processing method considering the coefficient matrix and observation vector containing both random errors and systematic errors is adopted.The examples show that the systematic errors of the bathymetry direction can be separated from the multibeam echosounding data and achieve certain effect.