Anisotropic Parameter Modeling Based on BP Neural Network

• DOI: 10.3997/2214-4609.202112466
• Publication Date: 2021-09-29T06:06:02Z
• Authors: S. Zhang; F. Wang; J. Huang; Z. Li
• Keywords: Snapshot (computer storage); Backpropagation; Data set

Summary In seismic data processing, anisotropic parameter modeling plays an important role in the migration imaging results. Traditional methods rely on a large amount of expert knowledge and have poor applicability. this paper, wave field snapshot set is constructed method based BP neural network proposed. First, homogeneous velocity model used to construct through numerical simulation. Then, carry out regression distribution. The experimental results show that has good applicability great application potential prediction parameters with error distribution convergence between [-0.001, 0.001].