基于ARX模型的地震资料提频方法
|
摘要
地层对地震波的吸收效应是限制地震数据分辨率的主要原因之一,其机制、估计与补偿对提高地震资料的分辨率具有重要意义。文中基于ARX模型,针对地层吸收过程进行逆向建模,将实际地面地震资料与井资料、井间地震资料分别作为模型的输入与输出,从而得到模型结构参数,建立地层吸收过程的逆向模型以进行高频信息补偿,达到地震资料的提频目的。对实际地震资料的处理结果表明,该方法可将主频提高10~15 Hz,频带拓宽8~10 Hz。在保侍原资料基本特征的同时,有效提高了地面地震资料的纵向分辨率。
Absorptive attenuation has always been one of the reasons for the limited resolution of the seismic data.The mechanism of the absorption in strata,the evaluation of the attenuation quantity and the compensation of the high frequency components are essential in the seismic resolution improvement.Based on ARX model,the authors simulated the reversion of the absorption process in strata by establishing a numerical model with system identification technology.The low frequency data(the surface seismic data) and the high frequency data(the well log data or the cross-well seismic data) are respectively set as the input and the output of the model,and then the structure parameters of the model can be obtained.The model in this paper can establish an interconnection between the conventional seismic data and the well log data,the cross-well data or the VSP data.It is shown by processing the practical data that the main frequency can be increased by 10 ~ 15 Hz,and the frequency band width can be expanded by 8 ~ 10 Hz.
Absorptive attenuation has always been one of the reasons for the limited resolution of the seismic data.The mechanism of the absorption in strata,the evaluation of the attenuation quantity and the compensation of the high frequency components are essential in the seismic resolution improvement.Based on ARX model,the authors simulated the reversion of the absorption process in strata by establishing a numerical model with system identification technology.The low frequency data(the surface seismic data) and the high frequency data(the well log data or the cross-well seismic data) are respectively set as the input and the output of the model,and then the structure parameters of the model can be obtained.The model in this paper can establish an interconnection between the conventional seismic data and the well log data,the cross-well data or the VSP data.It is shown by processing the practical data that the main frequency can be increased by 10 ~ 15 Hz,and the frequency band width can be expanded by 8 ~ 10 Hz.
引文
[1]Chopra S,Alexeev V,Sudhakar V.High-frequency restoration of surface seismic data[J].The Leading Edge,2003,22(8):730-738.
[2]曾锐,刘洪.柯西约束盲反褶积技术在井间地震的应用[J].地球物理学进展,2004,19(1):166-172.
[3]轩仁飞,曹俊兴.地震分倒频处理技术[J].地球物理学进展,2012,27(3):1183-1188.
[4]刘浩杰,王延光,韩文功.基于系统辨识提高地震资料分辨率研究[J].地球物理学进展,2010,25(3):1-5.
[5]李庆忠,王建花.井间地震勘探的误区及出路[J].石油地球物理勘探,2004,39(1):1-11.
[6]Bouzrara K,Garna T,Ragot J,et al.Online identification of the ARX model expansion on Laguerre orthonormal based with filters on model input and output[J].International Journal of Control,2013,86(3):369-385.
[7]刘文岭,夏海英.同位协同克里金方法在储层横向预测中的应用[J].勘探地球物理进展,2004,27(5):367-370.
[8]Gunnarsson S.Frequency domain accuracy of recursively identified ARX models[J].International Journal of Control,1991,54(2):465-480.
[9]周彤.面向控制的系统辨识导论[M].北京:清华大学出版社,2009.
[10]Millnert M.Identification of ARX models with markovian parameters[J].International Journal of Control,1987,45(6):2045-2058.
[11]Byehan S,Alci M.Fuzzy functions based ARX model and new fuzzy basis function models for nonlinear system identification[J].Applied Soft Computing Journal,2009,10(2):439-444.
[12]靳国栋,刘衍聪,牛文杰.距离加权反比插值法和克里金插值法的比较[J].长春工业大学学报:自然科学版,2003,31(3):46-48.
[13]Nounou M N,Nounou H N.Improving the prediction and parsimony of ARX models using multiscale estimation[J].Applied Soft Computing Journal,2006,7(3):711-721.
[14]Nicolao G D,Magni L,Scattolini R.Stabilizing predictive control of nonlinear ARX models[J].Automatica,1997,33(9):1691-1697.
[15]Golob M,Tovornik B.Input-output modeling with decomposed neuro-fuzzy ARX model[J].Neurocomputing,2007,71(4):875-884.
[2]曾锐,刘洪.柯西约束盲反褶积技术在井间地震的应用[J].地球物理学进展,2004,19(1):166-172.
[3]轩仁飞,曹俊兴.地震分倒频处理技术[J].地球物理学进展,2012,27(3):1183-1188.
[4]刘浩杰,王延光,韩文功.基于系统辨识提高地震资料分辨率研究[J].地球物理学进展,2010,25(3):1-5.
[5]李庆忠,王建花.井间地震勘探的误区及出路[J].石油地球物理勘探,2004,39(1):1-11.
[6]Bouzrara K,Garna T,Ragot J,et al.Online identification of the ARX model expansion on Laguerre orthonormal based with filters on model input and output[J].International Journal of Control,2013,86(3):369-385.
[7]刘文岭,夏海英.同位协同克里金方法在储层横向预测中的应用[J].勘探地球物理进展,2004,27(5):367-370.
[8]Gunnarsson S.Frequency domain accuracy of recursively identified ARX models[J].International Journal of Control,1991,54(2):465-480.
[9]周彤.面向控制的系统辨识导论[M].北京:清华大学出版社,2009.
[10]Millnert M.Identification of ARX models with markovian parameters[J].International Journal of Control,1987,45(6):2045-2058.
[11]Byehan S,Alci M.Fuzzy functions based ARX model and new fuzzy basis function models for nonlinear system identification[J].Applied Soft Computing Journal,2009,10(2):439-444.
[12]靳国栋,刘衍聪,牛文杰.距离加权反比插值法和克里金插值法的比较[J].长春工业大学学报:自然科学版,2003,31(3):46-48.
[13]Nounou M N,Nounou H N.Improving the prediction and parsimony of ARX models using multiscale estimation[J].Applied Soft Computing Journal,2006,7(3):711-721.
[14]Nicolao G D,Magni L,Scattolini R.Stabilizing predictive control of nonlinear ARX models[J].Automatica,1997,33(9):1691-1697.
[15]Golob M,Tovornik B.Input-output modeling with decomposed neuro-fuzzy ARX model[J].Neurocomputing,2007,71(4):875-884.
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