| In the process of measurement data,the problems people usually encounter are much more complicated than those expressed by the established function model.How to make full use of prior information to obtain reliable parameter information is a problem that measurement data processing must face.Due to the distance constraint,displacement and other prior information in surveying and mapping itself is nonlinear,it is more appropriate to describe them by using nonlinear constraint,and it is easier to integrate with adjustment criterion,but direct linearization of the constraint will make the constraint lose its significance.In this paper,the prior information is described in the form of nonlinear constraints and corresponding adjustment criteria are constructed according to its characteristics.Then,a new nonlinear constraint adjustment algorithm is established by extended regularization theory and applied to data processing in surveying and mapping.The main research contents are as follows:(1)Based on the characteristic that ridge parameters are related to constraints,a new method to determine ridge parameters is proposed,and the feasibility of the method is proved,and it is applied to solve ill-posed problems in measurement data processing.(2)In order to solve the problem of rank deficit in the measurement data processing,this paper uses a norm distance to treat the rank deficit matrix as a sick matrix with "almost rank deficit".The rank deficit adjustment model is transformed into a bounded prior information constrained adjustment model,and a new iterative method for solving the adjustment model is derived.(3)Data processing problems related to ellipse fitting in engineering measurement,such as section detection in tunnel engineering,need to measure the points on the component to fit the shape parameters before and after the ellipse deformation.The acquisition of measurement data points involves the conversion between source coordinates and target coordinates,which requires the invariance of ellipse fitting algorithm under coordinate transformation.In this paper,a new ellipse fitting algorithm with nonlinear constraints is presented.The results of numerical simulation show that the ellipse fitting algorithm is invariant under coordinate transformation. |
