基于U-Net神经网络构建三维磁基准图方法研究OA

The downward continuation algorithm of geomagnetic potential field is a typical ill-posed prob-lem in mathematics.Aimed at the problems that the traditional downward continuation of geomagnetic da-ta is susceptible to the effects of noise and continuation distance,this paper explores the downward contin-uation of magnetic anomalies by using the U-Net neural network.Magnetic anomalies obtained at different heightsvia forward modeling of simple prisms and complex combined prisms,are taken to be labels of ma-chine learning with 5 000 label groups being constructed for each dataset.The dataset is divided into train-ing,test,and validation sets at a ratio of 8∶1∶1.The construction of machine learning labels is in com-prehensive consideration information such as geometric position of the model center,the length,width,and height of the model,as well as magnetic inclination,magnetic declination,and magnetization intensi-ty.Under condition of interference-free experiments,the average relative prediction accuracy is 97.0％for the single prism model and 95.2％for the three-combined prism model.Under the interference of 5％Gaussian noise,the average relative prediction accuracy of the combined model remains 94.9％,showing excellent anti-interference performance.And simultaneously,the model exhibits strong distance robust-ness within the range of continuation distances involved in the experiments.This study verifies that the U-Net network can effectively realize downward continuation and has broad application prospects in the con-struction of 3D magnetic reference maps.