HARDY FUNCTION MODELING ELEVATION ANOMALY BASED ON TIKHONOV REGULARIZATION

In geoscience research, elevation anomaly modeling plays vital roles in engineering design and construction. In this paper, an improved Hardy function based on Tikhonov regularization is proposed to model elevation anomalies in which the improved model is not constrained by the ill -posed problem induced by that the number of nodes is less than the number of data points, and it has a wider range of use scenarios. In the experimental part, numerical simulation is used to generate simulated data of elevation anomaly, which is divided into training group and checking group, and the hardy functions before and after improvement are used to establish elevation anomaly models, and the performances of the two models are compared. is the results show that the improved Hardy function has better prediction performance with accuracy improvement of 79.75 % due to selecting more nodes and obtaining a more optimal model.