High-resolution Gr prediction method based on geological constraint and XGboost algorithm

The invention provides a high-resolution Gr prediction method based on geological constraints and an XGboost algorithm. The high-resolution Gr prediction method comprises the steps of 1, performing well seismic fine calibration; 2, aiming at a reservoir section, extracting a plurality of attributes on the seismic data; 3, in a time domain, seismic attributes and logging Gr values are extracted along well trajectories; step 4, seismic attribute optimization is carried out based on geological constraints; step 5, carrying out attribute optimization by using an XGboost algorithm; step 6, on the basis of attribute optimization, taking optimized attributes and logging Gr values as characteristic values and labels, and training a model by using XGboost; and step 7, predicting a Gr three-dimensional data volume. The high-resolution Gr prediction method based on the geological constraint and the XGboost algorithm is not affected by a low-frequency model, the inter-well reliability is high, the reliability degree is s