应用卡尔曼滤波识别微地震信号
|
摘要
本文在讨论卡尔曼滤波方法基础上,设计了微地震卡尔曼滤波的实现算法。在讨论裂缝产生微地震波传播的过程中,通过微地震波在地层中传播吸收衰减等因素影响分析,建立了适于微地震波传播的状态方程。针对微地震波传播过程中各种噪声的特点,研究了噪声的变化规律。就卡尔曼滤波中两类噪声方差初始估计问题,在假定噪声和信号不相关的条件下,研究了状态变量基于观测值无偏估计理论后,给出了观测初始噪声方差估计方法和与状态变量有关的输入噪声方差的估计方法。通过理论模型和实际资料进行处理、验证分析,证明方法可行。
Based on discussion of Kalman filtering, the algorithm to realize microseismic Kalman filtering was designed in this paper. During the discussion of microseismic wave propagation in fracture and through the analysis of affecting actors for microseismic wave propagation, Absorption and attenuation, the state equation which describes microseismic wave propagation was set up, and in accordance with the characteristics of all kinds of noise during the propagation process of microseismic wave the characteristic regularity for noise change were studied. Regarding two kinds of noise variance initial estimation problems in Kalman filtering, firstly supposing noise and signal not correlated with each other and then studying the state variables based on observed value unbiased estimation theory, at last the two estimation methods were obtained, the two methods are initial noise variance observation method and input noise variance method which is related to state variables. After processing and analyzing theoretical model and real data, the results show that the method is effective and woks well.
Based on discussion of Kalman filtering, the algorithm to realize microseismic Kalman filtering was designed in this paper. During the discussion of microseismic wave propagation in fracture and through the analysis of affecting actors for microseismic wave propagation, Absorption and attenuation, the state equation which describes microseismic wave propagation was set up, and in accordance with the characteristics of all kinds of noise during the propagation process of microseismic wave the characteristic regularity for noise change were studied. Regarding two kinds of noise variance initial estimation problems in Kalman filtering, firstly supposing noise and signal not correlated with each other and then studying the state variables based on observed value unbiased estimation theory, at last the two estimation methods were obtained, the two methods are initial noise variance observation method and input noise variance method which is related to state variables. After processing and analyzing theoretical model and real data, the results show that the method is effective and woks well.
引文
[1]邓自立.最优估计理论及其应用.黑龙江哈尔滨:哈尔滨工业大学出版社,2005,98~160
[2]朱光明等译.垂直地震剖面.北京:石油工业出版社,1983
[3]Stork A I,Ito H.Source parameter scaling for smallearthquake observed at the Western Nagano 800 mdeepborehole,central Japan.The Bulletin ofthe Seismolog-ical Society of America,2004,94(5):1781~1794
[4]Raab F,Coyne D.Effect of Micro-seismic Noise on aLIGO(Interferometer.California Institute of Technol-ogy),2001
[5]赵鸿儒,孙进忠,唐文榜等.全波震相分析.北京:地震出版社,1991
[6]张山,刘清林等.微地震监测技术在油田开发中的应用.石油物探,2002,41(2):226~231
[7]宋维琪,孙英杰.微地震资料频域相干—时间域偏振滤波方法.石油地球物理勘探,2008,43(2):161~167
[2]朱光明等译.垂直地震剖面.北京:石油工业出版社,1983
[3]Stork A I,Ito H.Source parameter scaling for smallearthquake observed at the Western Nagano 800 mdeepborehole,central Japan.The Bulletin ofthe Seismolog-ical Society of America,2004,94(5):1781~1794
[4]Raab F,Coyne D.Effect of Micro-seismic Noise on aLIGO(Interferometer.California Institute of Technol-ogy),2001
[5]赵鸿儒,孙进忠,唐文榜等.全波震相分析.北京:地震出版社,1991
[6]张山,刘清林等.微地震监测技术在油田开发中的应用.石油物探,2002,41(2):226~231
[7]宋维琪,孙英杰.微地震资料频域相干—时间域偏振滤波方法.石油地球物理勘探,2008,43(2):161~167
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