基于模拟退火和遗传算法的剩余静校正方法研究
|
摘要
静校正是贯穿于复杂地表地区地震勘探资料采集和处理中的一项十分关键的基础性工作。当地震资料信噪比比较低或者剩余静校正量大于地震波周期的一半时,为防止出现“周波跳跃”现象,这时就必须用全局优化方法寻求解决剩余静校正问题。论文首先详细阐述了静校正方法的分类及剩余静校正的非线性特点,介绍了模拟退火算法和遗传算法的基本原理及特点;其次,研究了二维观测系统的几何位置关系并设计了优化算法的目标函数;随后,分析了模拟退火和退火遗传算法的参数选择方案,将两种算法应用到理论模型及实际地震资料,静校正后叠加剖面的分辨率明显提高,层位连续性明显加强;最后,总结了论文编写过程中自己对静校正问题及优化算法的一点认识和建议。论文的关键工作在于:为减少算法的计算量,消除“周波跳跃”现象,选用超长道集的互相关作为优化算法的目标函数;利用模拟退火一步法来解决剩余静校正问题;结合模拟退火算法的思想,对遗传算法的目标函数进行尺度变换,实现退火遗传算法剩余静校正。
In complex surface areas, residual static correction is a pivotal and basic work, which is essential to seismic data acquisition and processing. When the S/N ratio is low or the residual statics are larger than half of the wavelet period, residual statics must be resolved by optimization methods to avoid the“cycle-skips”phenomenon.
The paper firstly formulates classifications of static correction methods and the nonlinear characteristics of residual static correction. Then the basic principles and characteristics of the simulated annealing and genetic algorithm are introduced. Secondly, the geometry relations of 2D survey is studied and the objective functions of the optimization algorithm are designed. Subsequently, after analyzing the schemes of parameter selections, the simulated annealing and hybrid genetic algorithm are applied to theoretical models and practical seismic data, which makes both resolutions of the stacked profiles and continuity of the layers enhanced. Finally, a little sentiment about static correction and optimization algorithm is summarized during this period.
In this paper, super-trace crosscorrelation is used as objective function to avoid“cycle-skips”and reduce computations. Secondly, the“heat bath”algorithm is used to calculate statics. Thirdly, combined with simulated annealing algorithm, the fitness function is scaled to implement hybrid genetic algorithm for static correction. All of this three points are the key work of this paper.
In complex surface areas, residual static correction is a pivotal and basic work, which is essential to seismic data acquisition and processing. When the S/N ratio is low or the residual statics are larger than half of the wavelet period, residual statics must be resolved by optimization methods to avoid the“cycle-skips”phenomenon.
The paper firstly formulates classifications of static correction methods and the nonlinear characteristics of residual static correction. Then the basic principles and characteristics of the simulated annealing and genetic algorithm are introduced. Secondly, the geometry relations of 2D survey is studied and the objective functions of the optimization algorithm are designed. Subsequently, after analyzing the schemes of parameter selections, the simulated annealing and hybrid genetic algorithm are applied to theoretical models and practical seismic data, which makes both resolutions of the stacked profiles and continuity of the layers enhanced. Finally, a little sentiment about static correction and optimization algorithm is summarized during this period.
In this paper, super-trace crosscorrelation is used as objective function to avoid“cycle-skips”and reduce computations. Secondly, the“heat bath”algorithm is used to calculate statics. Thirdly, combined with simulated annealing algorithm, the fitness function is scaled to implement hybrid genetic algorithm for static correction. All of this three points are the key work of this paper.
引文
[1] 董泽斌,吴良元.地震资料数据处理[M].石油工业出版社,1992: 33～45.
[2] 张新立,谢卫民等.柴达木盆地折射静校正方法的应用研究[J].石 油地球物理勘探,2002;37(B09):52～57.
[3] 刘先荣.折射波静校正技术在黄土塬地区的应用.内蒙古石油化工[J].2005;31(6):116～117.
[4] 郑军.变速场高程静校正在复杂地表中的应用[J].江汉石油职工大学学报,2004;17(4):4～6.
[5] 杨文军,段云卿,姜伟才,于富文.层析反演静校正[J].物探与化探,2005;29(1):41～43.
[6] Sheriff, R.E. Encyclopedic dictionary of exploration geophysics[M]. Soc. Expl. Geophys, 1991.
[7] 郭晓滨,李刚,冯永强.“三步法”静校正在鄂北麻黄山地区地震资料处理中的应用[J].海洋地质动态,2006;22(4):34～36.
[8] 戴云,张建中.长波长静校正问题的一种解决方法[J].石油地球物理勘探,2000;35(3):315～325.
[9] 车建英,张伟.初至折射波静校正技术在复杂探区地震数据处理中的应用[J].中国煤田地质,2006;18(1):54～56.
[10]李振春,张军华.地震数据处理方法[M].中国石油大学出版社,2004:146～181.
[11]Mike Cox 著.反射地震勘探静校正技术[M].李培明,柯本喜等译.石油工业出版社,2004.
[12]Taner M T, Koehler F and Alhilali K A. Estimation and correction of near-surface time anomalies[J]. Geophysics, 1974; 39(4): 441~463.
[13]Wiggins R, Larner K and Wisecup D. Residual statics analysis as a general linear inverse problem[J]. Geophysics, 1976; 41(5): 922~938.
[14]王振华,袁明生等.复杂地表条件下的静校正方法[J].石油地球物理勘探,2003;38(5):487～500.
[15]井西利.大剩余静校正量求解的两步法[J].石油地球物理勘探,2003;38(1):22～26.
[16]邢文训,谢金星.现代优化计算方法[M].清华大学出版社,1999:1～192.
[17]康立山,谢云.非数值并行算法(第一册)——模拟退火算法[M].科学出版社,1994.
[18]Szu H, Hartley R. Fast simulated annealing[J]. Phys Lett A, 1987;122: 157~162.
[19]Rothman D. H. Automatic estimation of large residual statics correction[J]. Geophysics, 1986; 51(2): 332~346.
[20]云庆夏,黄光球.遗传算法和遗传规划——种搜索寻优技术[M].冶金工业出版社,1997.
[21]王小平,曹立明.遗传算法——理论、应用与软件实现[M].西安交通大学出版社,2002.
[22]刘勇,康立山.非数值并行算法(第二册)——遗传算法[M].科学出版社,1995.
[23]陈国良.遗传算法及其应用[M].人民邮电出版社,1996.
[24]周明,孙树栋.遗传算法原理及应用[M].国防工业出版社,1999.
[25]陆基孟.地震勘探原理[M].石油大学出版社,1993.
[26]Claerbout, J. F., Black, J. L.. Basic Earth Imaging(Version 2.4)[M]. 2005.
[27]Ronen, J., Claertbout, J. F.. Surface-consistent residual statics estimation by stack-power maximization. Geophysics, 1985; 50(12): 2759~2767.
[28]Kirkpatrick, S., C.D. Gelatt and P.M. Vecchi. Optimization by Simulated Annealing[J]. Science, 1983; 220: 670~680.
[29]Rothman, D. H. Nonlinear inversion, statistical mechanics, andresidual statics estimation[J]. Geophysics, 1985; 50: 2784~2796.
[30]Landa, E., Beydoun, W. and Tarantola, A.. Reference velocity model estimation from prestack waveforms: Coherency optimization by simulated annealing[J]. Geophysics, 1989; 54: 984~990.
[31]Ammon, C. J. and Vidale, J, E.. Seismic traveltime tomography using combinatorial optimization techniques[J]. Seis. Res. Lett, 1990; 61: 39~40.
[32]Sen, M. K. and Stoffa, P. L.. Nonlinear one-dimensional seismic waveform inversion using simulated annealing[J]. Geophysics, 1991;56: 1624~1638.
[33]Sen, M. K., Bhattacharya, B. B. and Stoffa, P. L. Nonlinear inversion of resistivity sounding data[J]. Geophysics, 1993; 58: 496~507.
[34]Chunduru, R. K., Sen, M. K., Stoffa, P. L. et al. Nonlinear inversion of resistivity profiling data for some regular geometrical bodies[J].Geophys. Prosp, 1995; 43: 979~1003.
[35]Vasudevan, K., Wilson, W. G. and Laidlaw, W. G... Simulatedannealing statics computation using an order-based energy function[J]. Geophysics, 1991; 56(11): 1831~1839.
[36]John B., DuBose, Jr. Practical steps toward realizing the potential ofMonte Carlo automatic statics[J]. Geophysics, 1993; 58(3): 399~407.
[37]唐建侯,张金山.高效模拟退火剩余静校正[J].石油地球物理勘探,1994;29(3):382～387.
[38]刘鹏程,纪晨.改进的模拟退火单纯形综合反演方法[J].地球物理学报,1995;38(2):199～204.
[39]张红梅,吕小林等.改进的约束单纯形—模拟退火剩余静校正[J].地球物理学进展,2004;19(2):341～347.
[40]崔若飞,王勇.模拟退火剩余静校正的一步算法[J].物探化探计算技术,1996;18(3):242～247.
[41]尹成等.热槽法模拟退火分析及其改进[J].石油物探,1998;37(1):63～70.
[42]林依华,尹成等.一种新的求解静校正的全局快速寻优法[J].石油地球物理勘探,2000;35(1):1～12.
[43]林依华,张中杰等.复杂地形条件下静校正的综合寻优[J].地球物理学报,2003;46(1):101～106.
[44]姚付生,巫盛洪,杨旭明,彭怀春.混合优化算法剩余静校正[J].石油地球物理勘探,2002;37(增刊):71～74.
[45]苟量,彭真明等.混合优化自动剩余静校正方法研究[J].石油地球物理勘探,2005;40(4):454～458.
[46]曲凤山,井西利等.模拟退火剩余静校算(ARSCSA)及其并行实现[J].计算机工程,2001;27(5):119～122.
[47]井西利.大剩余静校正量求解的两步法[J].石油地球物理勘探,2003;38(1):22～26.
[48]井西利,杨长春等.地球物理问题的自适应模拟退火方法求解[J].地球物理进展,2003;18(2):327～330.
[49]井西利.模拟退火方法的改进及其在地震剩余静校正中的应用[J].西安石油学院学报,2003;18(4):19～22.
[50]井西利,李丽等.自适应混合反演剩余静校正[J].石油地球物理勘探,2004;39(2):153～157.
[51]井西利,杨长春.求解地震静校正问题的双尺度反演方法[J].地球物理学报,2006;49(5):1460～1466.
[52]D. S. Johnson, C. R. Aragon, L. A. McGeoch et al. Optimization by simulated annealing: An experimental evaluation, Part I[M]. AT&T Bell Laboratories, Murray Hill(NJ), 1987.
[53]P. J. M. Van Laarhoven, E. H. L. Aarts. Simulated annealing: Theory and applications[M]. D. Reidel, 1987.
[54]E. H. L. Aarts, J. H. M. Korst. Simulated annealing and Boltzmann machines[M]. John Wiley and Sons, 1989.
[55]Ingber, L. Very fast simulated annealing[J]. Math. Com. Modeling, 1989; 12: 967~993.
[56]Geman, D., Geman, S. Stochastic relaxation, Gibbs distributions and the Bayesian restoration of images[J]. Inst. Elec. Electron. Eng. Trans. On pattern analysis machine intelligence, 1984; PAM1-6, 721~741.
[57]赵军才,李铮等.塔中沙漠区地震资料静校正和去噪[J].勘探地球物理进展,2004;27(2):123～127.
[58]J. H. Holland. Adaptation in Natural and Artificial Systems[J]. University of Michigan Press, 1975.
[59]Stoffa, P. L. and Sen, M. K.. Nonlinear multiparameter optimization using genetic algorithms: Inversion of plane-wave seismograms[J].Geophysics, 1991; 56: 1794~1810.
[60]Wilson, W. G. and K. Vasudevan. Application of the genetic algorithm to residual statics estimation[J]. Geophys. Res. Lett., 1991; 18(12): 2181~2184.
[61]Sambridge, M. and Drijkoningen, G... Genetic algorithms in seismic waveform inversion[J]. Geophys. J. Int., 1992; 109: 323~342.
[62]Jin, S. and Madariaga, R. Background velocity inversion with agenetic algorithm[J]. Geophys. Res. Lett., 1993; 20: 93~96.
[63]Mathias, K., Whitley, D., Stork, C. and Kusuma, T. Staged hybridgenetic algorithm search for seismic data imaging[J]. Presented at theInternat. Conf. on Evolutionary Computing, ICEC 1994.
[64]William G. Wilson, William G. Laidlaw and Kris Vasudevan. Residual statics estimation using the genetic algorithm[J]. Geophysics, 1994;59(5): 766~774.
[65]Stork, C. and Kusuma, T. Hybrid Genetic Autostatitcs: New approach for large amplitude statics with noisy data[J]. Proceedings of the 62nd SEG meeting, 1993; New Orleans, OK, 1127~1131.
[66]Wenceslau Gouveia. Residual statics estimation by a distributed genetic algorithm[J]. SEG Technical Program Expanded Abstracts, 1994: 195~199.
[67]姚姚.用非线性遗传反演方法进行自动静校正[J].中国地质大学学报,1996;21(1):89～92.
[68]尹成.综合的并行搜索策略及其在剩余静校正中的应用[J].成都理工学院学报,1997;24(2):75～80.
[69]尹成.一种改进的遗传算法及其在剩余静校正中的应用[J].石油地球物理勘探,1997;32(4):486～491.
[70]李辉峰,邓飞,周熙襄.利用 TS 与 GA 的混合算法(TSGA )求取剩余静校正量[J].石油地球物理勘探,2006;41(3):327～332.
[2] 张新立,谢卫民等.柴达木盆地折射静校正方法的应用研究[J].石 油地球物理勘探,2002;37(B09):52～57.
[3] 刘先荣.折射波静校正技术在黄土塬地区的应用.内蒙古石油化工[J].2005;31(6):116～117.
[4] 郑军.变速场高程静校正在复杂地表中的应用[J].江汉石油职工大学学报,2004;17(4):4～6.
[5] 杨文军,段云卿,姜伟才,于富文.层析反演静校正[J].物探与化探,2005;29(1):41～43.
[6] Sheriff, R.E. Encyclopedic dictionary of exploration geophysics[M]. Soc. Expl. Geophys, 1991.
[7] 郭晓滨,李刚,冯永强.“三步法”静校正在鄂北麻黄山地区地震资料处理中的应用[J].海洋地质动态,2006;22(4):34～36.
[8] 戴云,张建中.长波长静校正问题的一种解决方法[J].石油地球物理勘探,2000;35(3):315～325.
[9] 车建英,张伟.初至折射波静校正技术在复杂探区地震数据处理中的应用[J].中国煤田地质,2006;18(1):54～56.
[10]李振春,张军华.地震数据处理方法[M].中国石油大学出版社,2004:146～181.
[11]Mike Cox 著.反射地震勘探静校正技术[M].李培明,柯本喜等译.石油工业出版社,2004.
[12]Taner M T, Koehler F and Alhilali K A. Estimation and correction of near-surface time anomalies[J]. Geophysics, 1974; 39(4): 441~463.
[13]Wiggins R, Larner K and Wisecup D. Residual statics analysis as a general linear inverse problem[J]. Geophysics, 1976; 41(5): 922~938.
[14]王振华,袁明生等.复杂地表条件下的静校正方法[J].石油地球物理勘探,2003;38(5):487～500.
[15]井西利.大剩余静校正量求解的两步法[J].石油地球物理勘探,2003;38(1):22～26.
[16]邢文训,谢金星.现代优化计算方法[M].清华大学出版社,1999:1～192.
[17]康立山,谢云.非数值并行算法(第一册)——模拟退火算法[M].科学出版社,1994.
[18]Szu H, Hartley R. Fast simulated annealing[J]. Phys Lett A, 1987;122: 157~162.
[19]Rothman D. H. Automatic estimation of large residual statics correction[J]. Geophysics, 1986; 51(2): 332~346.
[20]云庆夏,黄光球.遗传算法和遗传规划——种搜索寻优技术[M].冶金工业出版社,1997.
[21]王小平,曹立明.遗传算法——理论、应用与软件实现[M].西安交通大学出版社,2002.
[22]刘勇,康立山.非数值并行算法(第二册)——遗传算法[M].科学出版社,1995.
[23]陈国良.遗传算法及其应用[M].人民邮电出版社,1996.
[24]周明,孙树栋.遗传算法原理及应用[M].国防工业出版社,1999.
[25]陆基孟.地震勘探原理[M].石油大学出版社,1993.
[26]Claerbout, J. F., Black, J. L.. Basic Earth Imaging(Version 2.4)[M]. 2005.
[27]Ronen, J., Claertbout, J. F.. Surface-consistent residual statics estimation by stack-power maximization. Geophysics, 1985; 50(12): 2759~2767.
[28]Kirkpatrick, S., C.D. Gelatt and P.M. Vecchi. Optimization by Simulated Annealing[J]. Science, 1983; 220: 670~680.
[29]Rothman, D. H. Nonlinear inversion, statistical mechanics, andresidual statics estimation[J]. Geophysics, 1985; 50: 2784~2796.
[30]Landa, E., Beydoun, W. and Tarantola, A.. Reference velocity model estimation from prestack waveforms: Coherency optimization by simulated annealing[J]. Geophysics, 1989; 54: 984~990.
[31]Ammon, C. J. and Vidale, J, E.. Seismic traveltime tomography using combinatorial optimization techniques[J]. Seis. Res. Lett, 1990; 61: 39~40.
[32]Sen, M. K. and Stoffa, P. L.. Nonlinear one-dimensional seismic waveform inversion using simulated annealing[J]. Geophysics, 1991;56: 1624~1638.
[33]Sen, M. K., Bhattacharya, B. B. and Stoffa, P. L. Nonlinear inversion of resistivity sounding data[J]. Geophysics, 1993; 58: 496~507.
[34]Chunduru, R. K., Sen, M. K., Stoffa, P. L. et al. Nonlinear inversion of resistivity profiling data for some regular geometrical bodies[J].Geophys. Prosp, 1995; 43: 979~1003.
[35]Vasudevan, K., Wilson, W. G. and Laidlaw, W. G... Simulatedannealing statics computation using an order-based energy function[J]. Geophysics, 1991; 56(11): 1831~1839.
[36]John B., DuBose, Jr. Practical steps toward realizing the potential ofMonte Carlo automatic statics[J]. Geophysics, 1993; 58(3): 399~407.
[37]唐建侯,张金山.高效模拟退火剩余静校正[J].石油地球物理勘探,1994;29(3):382～387.
[38]刘鹏程,纪晨.改进的模拟退火单纯形综合反演方法[J].地球物理学报,1995;38(2):199～204.
[39]张红梅,吕小林等.改进的约束单纯形—模拟退火剩余静校正[J].地球物理学进展,2004;19(2):341～347.
[40]崔若飞,王勇.模拟退火剩余静校正的一步算法[J].物探化探计算技术,1996;18(3):242～247.
[41]尹成等.热槽法模拟退火分析及其改进[J].石油物探,1998;37(1):63～70.
[42]林依华,尹成等.一种新的求解静校正的全局快速寻优法[J].石油地球物理勘探,2000;35(1):1～12.
[43]林依华,张中杰等.复杂地形条件下静校正的综合寻优[J].地球物理学报,2003;46(1):101～106.
[44]姚付生,巫盛洪,杨旭明,彭怀春.混合优化算法剩余静校正[J].石油地球物理勘探,2002;37(增刊):71～74.
[45]苟量,彭真明等.混合优化自动剩余静校正方法研究[J].石油地球物理勘探,2005;40(4):454～458.
[46]曲凤山,井西利等.模拟退火剩余静校算(ARSCSA)及其并行实现[J].计算机工程,2001;27(5):119～122.
[47]井西利.大剩余静校正量求解的两步法[J].石油地球物理勘探,2003;38(1):22～26.
[48]井西利,杨长春等.地球物理问题的自适应模拟退火方法求解[J].地球物理进展,2003;18(2):327～330.
[49]井西利.模拟退火方法的改进及其在地震剩余静校正中的应用[J].西安石油学院学报,2003;18(4):19～22.
[50]井西利,李丽等.自适应混合反演剩余静校正[J].石油地球物理勘探,2004;39(2):153～157.
[51]井西利,杨长春.求解地震静校正问题的双尺度反演方法[J].地球物理学报,2006;49(5):1460～1466.
[52]D. S. Johnson, C. R. Aragon, L. A. McGeoch et al. Optimization by simulated annealing: An experimental evaluation, Part I[M]. AT&T Bell Laboratories, Murray Hill(NJ), 1987.
[53]P. J. M. Van Laarhoven, E. H. L. Aarts. Simulated annealing: Theory and applications[M]. D. Reidel, 1987.
[54]E. H. L. Aarts, J. H. M. Korst. Simulated annealing and Boltzmann machines[M]. John Wiley and Sons, 1989.
[55]Ingber, L. Very fast simulated annealing[J]. Math. Com. Modeling, 1989; 12: 967~993.
[56]Geman, D., Geman, S. Stochastic relaxation, Gibbs distributions and the Bayesian restoration of images[J]. Inst. Elec. Electron. Eng. Trans. On pattern analysis machine intelligence, 1984; PAM1-6, 721~741.
[57]赵军才,李铮等.塔中沙漠区地震资料静校正和去噪[J].勘探地球物理进展,2004;27(2):123～127.
[58]J. H. Holland. Adaptation in Natural and Artificial Systems[J]. University of Michigan Press, 1975.
[59]Stoffa, P. L. and Sen, M. K.. Nonlinear multiparameter optimization using genetic algorithms: Inversion of plane-wave seismograms[J].Geophysics, 1991; 56: 1794~1810.
[60]Wilson, W. G. and K. Vasudevan. Application of the genetic algorithm to residual statics estimation[J]. Geophys. Res. Lett., 1991; 18(12): 2181~2184.
[61]Sambridge, M. and Drijkoningen, G... Genetic algorithms in seismic waveform inversion[J]. Geophys. J. Int., 1992; 109: 323~342.
[62]Jin, S. and Madariaga, R. Background velocity inversion with agenetic algorithm[J]. Geophys. Res. Lett., 1993; 20: 93~96.
[63]Mathias, K., Whitley, D., Stork, C. and Kusuma, T. Staged hybridgenetic algorithm search for seismic data imaging[J]. Presented at theInternat. Conf. on Evolutionary Computing, ICEC 1994.
[64]William G. Wilson, William G. Laidlaw and Kris Vasudevan. Residual statics estimation using the genetic algorithm[J]. Geophysics, 1994;59(5): 766~774.
[65]Stork, C. and Kusuma, T. Hybrid Genetic Autostatitcs: New approach for large amplitude statics with noisy data[J]. Proceedings of the 62nd SEG meeting, 1993; New Orleans, OK, 1127~1131.
[66]Wenceslau Gouveia. Residual statics estimation by a distributed genetic algorithm[J]. SEG Technical Program Expanded Abstracts, 1994: 195~199.
[67]姚姚.用非线性遗传反演方法进行自动静校正[J].中国地质大学学报,1996;21(1):89～92.
[68]尹成.综合的并行搜索策略及其在剩余静校正中的应用[J].成都理工学院学报,1997;24(2):75～80.
[69]尹成.一种改进的遗传算法及其在剩余静校正中的应用[J].石油地球物理勘探,1997;32(4):486～491.
[70]李辉峰,邓飞,周熙襄.利用 TS 与 GA 的混合算法(TSGA )求取剩余静校正量[J].石油地球物理勘探,2006;41(3):327～332.