Research on the Reconstruction Method of Geomagnetic Observation Data Based on Improved BP Neural Network

In response to the problem of large-scale data loss caused by power outages, instrument failures, and the removal of abnormal data during the observation of various components of geomagnetic data in Shandong Province, this paper proposes a geomagnetic observation data reconstruction method based on an improved BP neural network. By selecting a reference station within a certain range from the missing data station as input, the predicted output value is used as the reconstruction data for the missing data station, The application of this method to the interpolation processing of geomagnetic data in Shandong region has achieved good results, solving the problems of low effectiveness and poor universality of current interpolation reconstruction methods for geomagnetic observation data.