Sparse restoration of fractal self-similar data and its application in uranium distribution

Abstract

Uranium is a rare and important energy mineral. In order to study the distribution characteristics of uranium deposits, we establish a data restoration algorithm with fractal features and sparse known point data. The algorithm ensures the accuracy of the restoration and the retention of fractal features through the steps of determining the basis function, preserving the dimension invariant restoration, and dimension greedy optimization. By comparing the numerical test with the Kriging interpolation method, it is proved that the algorithm is better in accuracy error and dimension error when restoring the fractal feature data.