摘要
基于分形先验信息的非线性反演方法能综合利用测井数据和地震信息,在贝叶斯框架下,通过分形高斯噪音算法得到基于分形理论的先验信息,然后根据地震资料构建似然函数,最终利用基于快速模拟退火算法(Very Fast Simulated Annealing,VFSA)改进的粒子群优化(Particle Swarm Optimization,PSO)算法(VFSA-PSO)实现后验概率密度的抽样。与确定性反演结果相比,该方法能够有效地融合测井资料中的高频信息,提高反演结果的分辨率,并且目标函数的建立融合了确定性反演中的低频约束,从而得到宽频带的反演结果。数值模拟试验表明:基于分形先验的非线性反演结果与理论模型吻合较好,实际资料的应用效果也证明了该反演方法的有效性。
Nonlinear inversion based on fractal priori information is formulated in a Bayesian framework,combining logging data with seismic information.Firstly,we can get the priori information through fractal Gaussian noise algorithm.Then,based on the seismic information,we can construct the likelihood function.Finally we apply VFSA-PSO optimization algorithm to obtain an exhaustive characterization of the posteriori probability density.Compared to deterministic inversion,the method we proposed can integrate the high frequency information of well logging and have a higher resolution.And the establishment of the objective function adds constraints of low frequency information using the experience of deterministic inversion for reference,so the inversion results contain the broadband information.According to the numerical calculations,we can conclude that the nonlinear inversion results based on fractal priori information match the model well.The application of real data also proves the effectiveness of the proposed nonlinear inversion method.
引文
[1]FRANCIS A.Limitations of deterministic and advantages of stochastic seismic inversion[J].CSEG Recorder,2005:5-11.
[2]印兴耀,刘永社.储层建模中地质统计学整合地震数据的方法及研究进展[J].石油地球物理勘探,2002,37(4):423-430.YIN X Y,LIU Y S.Methods and development of integrating seismic data in reservoir model-building[J].Oil Geophysical Prospecting,2002,37(4):423-430.(In Chinese)
[3]SRIVASTAVA R P,SEN M K.Stochastic inversion of prestack seismic data using fractal-based initial models[J].Geophysics,2010,75(3):R47-R59.
[4]TAO Y,SPIKES K,SEN M K.Stochastic seismic inversion using both fractal and low-frequency priors[C]//2011SEG Annual Meeting.Society of Exploration Geophysicists,2011.
[5]KIRKPATRICK S.Optimization by simulated annealing:Quantitative studies[J].Journal of statistical physics,1984,34(5-6):975-986.
[6]JOHN H.Adaptation in natural and artificial systems[M].Holland:Ann Arbor University of Michigan Press,1992.
[7]张广智,王丹阳,印兴耀.利用MCMC方法估算地震参数[J].石油地球物理勘探,2011,46(4):605-609.ZHANG G Z,WANG D Y,YIN X Y.Seismic parameter estimation using Markov Chain Monte Carlo Method[J].Oil Geophysical Prospecting,2011,46(4):605-609.(In Chinese)
[8]KENNEDY J,EBERHART R.Particle swarm optimization[C].Neural Networks,1995IEEE International Conference on Proceedings.1995:1942-1948.
[9]EBERHART R C,KENNEDY J.A new optimizer using particle swarm theory[C]//Proceedings of the sixth international symposium on micro machine and human science,1995,1:39-43.
[10]高鹰,谢胜利.基于模拟退火的粒子群优化算法[J].计算机工程与应用,2004,40(1):47-50.GAO Y,XIF S L.Particle swarm optimization algorithms based on simulated annealing[J].Computer Engineering and Applications,2004,40(1):47-50.(In Chinese)
[11]KJSBERG H,HAUGE R,KOLBJRNSEN O,et al.Bayesian Monte Carlo method for seismic predrill prospect assessment[J].Geophysics,2010,75(2):O9-O19.
[12]JOHN A,SCALES,MARTIN L,et al.Introductory Geophysical Inverse Theory[M].Samizdat Press,2001.
[13]HEWETT T A.Fractal distributions of reservoir heterogeneity and their influence on fluid transport[C]//SPE Annual Technical Conference and Exhibition.Society of Petroleum Engineers,1986.
[14]韩忠玲.分形高斯噪声Hurst参数估计的实验评价[D].华东师范大学,2008.HAN Z L.Experimental evaluation on methods of the Hurst parameter estimation using fractional Gaussian noise[D].Shanghai:East China Normal University,2008.(In Chinese)
[15]朱丽莉,杨志鹏,袁华.粒子群优化算法分析及研究进展[J].计算机工程与应用,2007,43(5):24-27.ZHU L L,YANG Z P,YUAN H.Analysis and development of particle swarm optimization[J].Computer Engineering and Applications,2007,43(5):24-27.(In Chinese)
[16]ANGELINE P J.Evolutionary optimization versus particle swarm optimization:Philosophy and performance differences[C]//Evolutionary Programming VII.Springer Berlin Heidelberg,1998:601-610.
[17]HIGASHI N,IBA H.Particle swarm optimization with Gaussian mutation[C]//Swarm Intelligence Symposium,2003.SIS'03.Proceedings of the 2003IEEE.IEEE,2003:72-79.
[18]METROPOLIS N,ROSENBLUTH A W,ROSENBLUTH M N,et al.Equation of state calculations by fast computing machines[J].The journal of chemical physics,2004,21(6):1087-1092.
[19]ROTHMAN D H.Nonlinear inversion statistical mechanics and residual statics estimation[J].Geophysics,1985,50(12):2784-2796.
[20]ROTHMAN D H.Automatic estimation of large residual statics correction[J].Geophysics,1986,51(2):337-346.
[21]张广智,李宁,李超,等.基于非常快速模拟退火算法的地震谱反演方法[J].西北地震学报,2012,34(4):313-323.ZHANG G Z,LI N,LI C,et al.Spectral inversion based on very fast simulated annealing algorithm[J].Northwestern Seismological Journal,2012,34(4):313-323.(In Chinese)