Research On Inversion Of Optimization Method Based On Boundary Recognition Of Magnetic Gradient Tensor

Magnetic tensor exploration is a high-precision exploration method,which has the advantages of large amount of information and high resolution.At present,the application prospect of magnetic gradient tensor exploration is increasingly broad,and the research on tensor data processing methods has been widely concerned.Boundary recognition is an indispensable task in magnetic field data interpretation.The magnetic gradient tensor data has higher accuracy for directly delineating the boundary of the abnormal body and delineating the underground abnormal body.Based on the magnetic gradient tensor data,a variety of boundary recognition methods have been formed,which have been widely used in the processing and interpretation of magnetic anomalies.But in most cases,it is mainly used for qualitative explanation.The optimization inversion method in magnetic anomaly processing is a quantitative interpretation method,which can be divided into geometry inversion and physical property inversion.At present,regular fixed section model is mainly used for physical property inversion.When there is more prior information and the shape of three-dimensional magnetic body is similar to regular model,geometric parameter inversion can also be carried out.In this paper,we combine the identification of magnetic gradient tensor boundary with the inversion of magnetic anomaly optimization method,and analyze the inversion of magnetic anomaly optimization method based on the identification of magnetic gradient tensor boundary.In this paper,firstly,the principle and properties of the magnetic gradient tensor are briefly introduced,and the specific formula of the conversion of the magnetic gradient tensor under different conditions in the frequency domain is simply derived.The boundary recognition is carried out by using the components of the magnetic gradient tensor and the tensor invariants I1 and I2.Then,the principle of the optimal inversion of magnetic anomalies is described,the principle of the optimal selection method in the sense of least square is introduced,and the objective function form constructed by the optimization method in this paper is explained.Then,through model trial,two commonly used optimization methods,simplexmethod and quasi Newton method,are used for inversion,and the effect of different initial values,noises and magnetic parameter errors on the optimization inversion results are compared and analyzed,which is used to study the fault tolerance of the optimization method based on the boundary identification of magnetic gradient tensor,so as to apply the method in practice.Finally,this paper uses the method to process and explain the data of magnetic anomaly in mining area,and explores its effect in practical application.In this paper,the components of the magnetic gradient tensor and the tensor invariants I1 and I2 are synthetically used for boundary recognition.The qualitative interpretation results of the boundary recognition are taken as the initial value of the quantitative inversion of the boundary position.The geometric parameters are quantitatively inverted by the simplex method and the quasi Newton method,which are two commonly used methods for the optimization of the magnetic anomalies.Through the inversion of the optimization method,the geometric parameters of the magnetic body model can be determined more accurately,and the boundary of the magnetic body can be depicted.This paper explores the method of quantitative boundary recognition,which has certain theoretical and practical significance.