Simulation And Analysis Of Earth’s Magnetic Field Based On B-Spline Fitting

Based on the CHAMP satellite providing high accuracy,high density of scalar and vector global geomagnetic observation data,this article select the suitable curve fitting and surface fitting method and research the numerical analysis of magnetic data.Finally,we expected magnetic field intensity values are presented on a complete sphere.Based on the b-spline basic theories,respectively,using the method of least squares method,based on L-M and least squares algorithm and support vector machine method to collect data for curve and surface fitting.The fitting performance and effect of the three algorithms are studied.Firstly,using least square fitting method of b-spline curve and surface fitting to research the problem,which involving mass equations.This method runs slow and fitting accuracy is also low.To solve these problems,the algorithm has been improved.Based on L,M method and hybrid algorithm of least square fitting method of b-spline curves and surface fitting,speed and accuracy are improved.Thus,the algorithm takes up memory and data size is large,so that the current computer can not satisfy the storage requirements.When kernel function is b-spline basis functions,using support vector machines algorithm for curve fitting,solving support vector machines can be converted to solve convex quadratic programming problems.It may be the only global optimal solution.In the simulation process,the support vector machine for b-spline fitting has advantage on accuracy and speed.Finally,the fitted closed spherical surface fits the requirements of the subject.