Penalized total least squares method for dealing with systematic errors in partial EIV model and its precision estimation

When the total least squares(TLS)solution is used to solve the parameters in the errors-in-variables(EIV)model,the obtained parameter estimations will be unreliable in the observations containing systematic errors.To solve this problem,we propose to add the nonparametric part(systematic errors)to the partial EIV model,and build the partial EIV model to weaken the influence of systematic errors.Then,having rewritten the model as a nonlinear model,we derive the formula of parameter estimations based on the penalized total least squares criterion.Furthermore,based on the second-order approximation method of precision estimation,we derive the second-order bias and covariance of parameter estimations and calculate the mean square error(MSE).Aiming at the selection of the smoothing factor,we propose to use the U curve method.The experiments show that the proposed method can mitigate the influence of systematic errors to a certain extent compared with the traditional method and get more reliable parameter estimations and its precision information,which validates the feasibility and effectiveness of the proposed method.

An improved mixed total least squares method for strain inversion from distance changes

Based on the deficiency of the traditional total least squares method（TLS） in the field of geodetic inversion, the mixed error characteristics of the errors in variables（EIV） model were analyzed by considering the distance azimuth measurement error in strain inversion from distance changes, which resulted in the improved mixed total least squares method（IMTLS） with generalized mixed EIV model. Finally, three comparison schemes of strain inversion from distance changes using measured data were implemented to test the proposed method. The results showed that the IMTLS method outperformed the traditional least squares（LS） and TLS methods in parameters estimation, accuracy evaluation, actual EIV model characteristics, and five strain eigenvalues.

Robust estimation algorithm for multiple-structural data

This paper proposes a robust method of parameter estimation and data classification for multiple-structural data based on the linear error in variable（EIV） model.The traditional EIV model fitting problem is analyzed and a robust growing algorithm is developed to extract the underlying linear structure of the observed data.Under the structural density assumption,the C-step technique borrowed from the Rousseeuw＇s robust MCD estimator is used to keep the algorithm robust and the mean-shift algorithm is adopted to ensure a good initialization.To eliminate the model ambiguities of the multiple-structural data,statistical hypotheses tests are used to refine the data classification and improve the accuracy of the model parameter estimation.Experiments show that the efficiency and robustness of the proposed algorithm.