Chen, Guang (China University of Petroleum–Beijing) | Zhou, Hui (China University of Petroleum–Beijing) | Liu, Mingdi (China University of Petroleum–Beijing) | Tao, Yonghui (China University of Petroleum–Beijing) | Wang, Haiyang (China University of Petroleum–Beijing)
Elastic impedance inversion is an important prestack inversion method in reservoir prediction and fluid identification. Constrained sparse spike inversion (CSSI) used to be the most widely used method in poststack seismic inversion. Because of the relationship between the elastic impedance and the wave impedance, the CSSI can be directly applied to the prestack elastic impedance inversion. However, CSSI usually suffers from strongly ill-posed problem when using some local optimization algorithm, such as conjugate gradient (CG) method, and has a strong dependence on the initial model of reflection coefficient. Besides, conventional CSSI separately inverts the time locations and amplitudes of sparse-spikes, which increases the computational complexity. In this paper, we improve CSSI theory with a linear wave impedance constraint, which is named as LCSSI. We apply orthogonal matching pursuit (OMP) non-linear algorithm to linear constrained sparse spike prestack elastic impedance inversion. We derive a linear matrix equation from the cost function and use OMP algorithm to invert the sparse-spikes' time locations and amplitudes simultaneously. Due to OMP, this method can reduce dependence on initial model and obtain the inversion results precisely and quickly. Numerical examples indicate that our method can achieve a good performance even for noisy data, while the CG algorithm fails to get a desirable inversion result.
Presentation Date: Wednesday, October 19, 2016
Start Time: 8:00:00 AM
Presentation Type: ORAL