Nondependent-derivative method to process nonlinear data in digital science engineering
(Department of Territory Information and Surveying Engineering, Xuzhou Normal Universit, Xuzhou 221009, China)
Abstract: Data, including the spatial data and the non-spatial data, are the basis of all digital scientific engineering projects, such as the digital earth and the digital nation, the digital mine. The spatial data have the characteristics of many sources, multi-dimension, multi-type, many time states and different accuracy. The spatial data firstly must be processed before using these data. The parameter estimation model to process the data is commonly the more complex nonlinear model including random parameters and non-random parameters. So a generalized nonlinear dynamic least squares method to process these data is put forward. According to the special structure of the generalized nonlinear dynamic least squares problem and the solution to the first order, a new solving model and a corresponding method to process the problem are put forward. The complex problem can be divided into two sub-problems so that the number of the unknown parameters is reduced largely. Therefore it reduces the computing difficulty and load.
Key words: Generalized nonlinear dynamic least squares method； Separating algorithm； Difference quotient