Multiparameter prestack seismic inversion is one of the most powerful techniques in quantitatively estimating subsurface petrophysical properties. However, it remains a challenging problem due to the nonlinearity and ill-posedness of the inversion process. Traditional regularization approach can stabilize the solution but at the cost of smoothing valuable geological boundaries. In addition, compared with linearized optimization methods, global optimization techniques can obtain better results regardless of initial models, especially for multiparameter prestack inversion. However, when solving multiparameter prestack inversion problems, the application of standard global optimization algorithms maybe limited due to the issue of high computational cost (e.g., simulating annealing) or premature convergence (e.g., particle swarm optimization). In this paper, we propose a hybrid optimization-based multiparameter prestack inversion method. In this method, we introduce a prior constraint term featured by multiple regularization functions, intended to preserve layered boundaries of geological formations; in particular, to address the problem of premature convergence existing in standard particle swarm optimization algorithm, we propose a hybrid optimization strategy by hybridizing particle swarm optimization and very fast simulating annealing to solve the nonlinear optimization problem. We demonstrate the effectiveness of the proposed inversion method by conducting synthetic test and field data application, both of which show encouraging results.
This is a preview of subscription content, log in to check access.
The authors gratefully acknowledge the National Natural Science Foundation of China (41674113) (41374116), the Project of CNOOC (CNOOC-KJ-125-ZDXM-07-LTD-NFGC-2014-04), and the Fundamental Research Funds for the Central Universities (KYLX16_0758) for supporting the research. The authors appreciate the two anonymous reviews for offering constructive comments leading to great improvements in this paper.
Aki K, Richards PG (1980) Quantitative seismology: theory and methods. W.H. Freeman and Company, San FranciscoGoogle Scholar