约束模拟退火反演
|
摘要
模拟退火算法的最大优点是可以突破寻优过程中局部最优的限制，所以该方法适用于解决地震反演中普遍存在的非凸性目标函数的最优化问题。本文讨论了用模拟退火（SA）法进行地震资料的参数又演以及约束条件对反演过程的控制作用。约束条件包括模型参数取慎范围和选择初始模型，加入约束条件提高了模拟退火反演的收敛速度，使反演解更为合理。通过理论模型试算和实际资料试处理，证明该方法精度高、实用性强。
The most important superiority of simulative annealing method is to breakthrough the Iimitation of locaI optimization in an optirnization process,therefore themethod is applicable to the optimization of the nonwconvex objective function that iswideIy existent in seismic inversion. The article describes both the seismic parame-ter inversion using simulative annealing method and the influence of restrictive con-dition upon inversion process- Restrictive condition includes the range of tnodel pa-rameters and the chosen initial model. Restrictive condition speeds up the conver-gence of simulative annealing inversion to rnake the inversion solution more reason-able- Theoretical modeling and reaI data processing prove the method accurate andpractical.
The most important superiority of simulative annealing method is to breakthrough the Iimitation of locaI optimization in an optirnization process,therefore themethod is applicable to the optimization of the nonwconvex objective function that iswideIy existent in seismic inversion. The article describes both the seismic parame-ter inversion using simulative annealing method and the influence of restrictive con-dition upon inversion process- Restrictive condition includes the range of tnodel pa-rameters and the chosen initial model. Restrictive condition speeds up the conver-gence of simulative annealing inversion to rnake the inversion solution more reason-able- Theoretical modeling and reaI data processing prove the method accurate andpractical.
引文
1Gauthier O,Virieux J,Taratola A. Two-dimensional nonlimear inversion of seismic waveforms:Numericalresults.Geophysics,1986,51(7):1387～1403
2Metropolis N et al Equation of state calculations by fast computing machines.J Chem Phys,1953,21:1087～1092
3 Kirkpatrick S,et al. Optimization by simulated annealing. Science, 1983, 220: 671 ~ 680
4 Rothman D H. NonIinear inversion l statistical mechanics,and residual statics estimation. GeoPhysics, 1985,50(12): 2784~ 2796
5 Rothman D H. Automatic estimation of large residual statics correction. GeoPhysics, 1986, 51 (2): 332~ 346
6 Sen M K,Stoffa P L. Nonlinear one-dimensional seismic wave form inversion using simuIated annealing.GeoPhysics, 1991, 56 (10): 1624~1638
7 Martinez R D et al- Complex reservoir characterization by mu1tiparameter constrainted inversion. Present-ed at the SEG/EAEG Research Workshop on Reservoir Geophysics, 1988
8 Mansfield G. Interactive broadband constrainted inversion. Eoploration GeoPhysics, 1989, 20 (1-2 ): 253 ~255
9 Vest R T.利用地质约束条件进行地震数据反演,第 60届SEG年会论文集,石油工业出版社,1992
10苏汝铿.统计物理学,复旦大学出版社,1990
11 Geman S, Geman D. Stochastic relation, Gibbs distribution and Bayesian restoration of images. IEEETrans, 1984, PAMI-6 (6): 721 ~ 741
12王梓坤.概率论基础及其应用,科学出版社,1979
2Metropolis N et al Equation of state calculations by fast computing machines.J Chem Phys,1953,21:1087～1092
3 Kirkpatrick S,et al. Optimization by simulated annealing. Science, 1983, 220: 671 ~ 680
4 Rothman D H. NonIinear inversion l statistical mechanics,and residual statics estimation. GeoPhysics, 1985,50(12): 2784~ 2796
5 Rothman D H. Automatic estimation of large residual statics correction. GeoPhysics, 1986, 51 (2): 332~ 346
6 Sen M K,Stoffa P L. Nonlinear one-dimensional seismic wave form inversion using simuIated annealing.GeoPhysics, 1991, 56 (10): 1624~1638
7 Martinez R D et al- Complex reservoir characterization by mu1tiparameter constrainted inversion. Present-ed at the SEG/EAEG Research Workshop on Reservoir Geophysics, 1988
8 Mansfield G. Interactive broadband constrainted inversion. Eoploration GeoPhysics, 1989, 20 (1-2 ): 253 ~255
9 Vest R T.利用地质约束条件进行地震数据反演,第 60届SEG年会论文集,石油工业出版社,1992
10苏汝铿.统计物理学,复旦大学出版社,1990
11 Geman S, Geman D. Stochastic relation, Gibbs distribution and Bayesian restoration of images. IEEETrans, 1984, PAMI-6 (6): 721 ~ 741
12王梓坤.概率论基础及其应用,科学出版社,1979
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心