Research on nonlinear inversion of seismic surface waves based on artificial neural network algorithm
WANG Yiming1, SONG Xianhai1,2, ZHANG Xueqiang1
1. Institute of Geophysics & Geomatics, China University of Geosciences, Wuhan, Hubei 430074, China; 2. Hubei Subsurface Multi-scale Imaging Lab(SMIL), China University of Geosciences, Wuhan, Hubei 430074, China
Abstract：Inversion of S-wave velocity in underground media by Rayleigh wave dispersion curves is an effective and reliable geophysical method that has been widely used in the near-surface geological survey. Traditional linear inversion cannot meet the needs of increasingly complex targets or tasks, while nonlinear inversion has attracted much attention due to fast inversion speed and the intuitively understandable principle. The artificial neural network in nonlinear inversion has self-organizing and self-learning ability, associative memory, and strong fault-tolerance and anti-interference. For different inversion problems, it can output corresponding target parameters (by network training). In this regard, the BP artificial neural network was adopted in this paper to inverse seismic surface waves, and the optimal solution rather than exact solution was obtained by designing, training, and learning, which avoided the problem that other nonlinear inversion methods were easy to fall into local minima. The BP neural network was selected to invert the dispersion curves of typical geological models in noiseless and noisy environments, and the inversion results were compared with the real data to verify the effectiveness and stability of this method. Finally, the measured data was inverted and compared with the inversion results of other methods to verify the practicability of the proposed method.
王一鸣, 宋先海, 张学强. 应用人工神经网络算法的地震面波非线性反演[J]. 石油地球物理勘探, 2021, 56(5): 979-991.
WANG Yiming, SONG Xianhai, ZHANG Xueqiang. Research on nonlinear inversion of seismic surface waves based on artificial neural network algorithm. Oil Geophysical Prospecting, 2021, 56(5): 979-991.
邵广周, 李庆春.联合应用τ-p变换法和相移法提取面波频散曲线[J]. 石油地球物理勘探, 2010, 45(6):836-840.SHAO Guangzhou, LI Qingchun.Joint application of τ-p and phase-shift stacking method to extract ground wave dispersion curve[J]. Oil Geophysical Prospecting, 2010, 45(6):836-840.
[2]
鲁来玉, 张碧星, 汪承灏.基于瑞利波高阶模式反演的实验研究[J]. 地球物理学报, 2006, 49(4):1082-1091.LU Laiyu, ZHANG Bixing, WANG Chenghao.Experiment and inversion studies on Rayleigh wave considering higher modes[J]. Chinese Journal of Geophy-sics, 2006, 49(4):1082-1091.
[3]
宋先海, 张学强, 王一鸣, 等.近地表弹性介质瑞雷波勘探研究进展与展望[J]. 地质科技通报, 2020, 39(5):173-182.SONG Xianhai, ZHANG Xueqiang, WANG Yiming, et al.Recent advances and prospects of near surface elastic Rayleigh waves[J]. Bulletin of Geological Science and Technology, 2020, 39(5):173-182.
[4]
Xu Y X, Xia J H, Miller R D.Numerical investigation of implementation of air-earth boundary by acoustic-elastic boundary approach[J]. Geophysics, 2007, 72(5):SM147-SM153.
[5]
高静怀, 何洋洋, 马逸尘.黏弹性与弹性介质中Rayleigh面波特性对比研究[J]. 地球物理学报, 2012, 55(1):207-218.GAO Jinghuai, HE Yangyang, MA Yichen.Comparison of the Rayleigh wave in elastic and viscoelastic media[J]. Chinese Journal of Geophysics, 2012, 55(1):207-218.
[6]
夏江海, 高玲利, 潘雨迪, 等.高频面波方法的若干新进展[J]. 地球物理学报, 2015, 58(8):2591-2605.XIA Jianghai, GAO Lingli, PAN Yudi, et al.New fin-dings in high frequency surface wave method[J]. Chinese Journal of Geophysics, 2015, 58(8):2591-2605.
[7]
Chai H Y, Cui Y J, Wei C F.A parametric study of effective phase velocity of surface waves in layered media[J]. Computers & Geotechnics, 2012, 44(June):176-184.
[8]
Zhang S X, Chan L S, Chen C Y, et al.Apparent phase velocities and fundamental-mode phase velocities of Rayleigh waves[J]. Soil Dynamics and Earthquake Engineering, 2003, 23(7):563-569.
[9]
Li J, Lin F C, Allam A, et al.Wave equation dispersion inversion of surface waves recorded on irregular topography[J]. Geophysical Journal International, 2019, 217(1):346-360.
曹旭, 熊章强, 张大洲.基于BP神经网络的瑞雷面波智能优化反演[J]. 工程地球物理学报, 2015, 12(4):514-519.CAO Xu, XIONG Zhangqiang, ZHANG Dazhou.The Rayleigh surface wave intelligent inversion based on the BP artificial neural network[J]. Chinese Journal of Engineering Geophysics, 2015, 12(4):514-519.
[12]
于东凯, 宋先海, 张学强, 等.蚱蜢算法在瑞雷波频散曲线反演中的应用[J]. 石油地球物理勘探, 2019, 54(2):288-301.YU Dongkai, SONG Xianhai, ZHANG Xueqiang, et al.Rayleigh wave dispersion inversion based on grasshopper optimization algorithm[J]. Oil Geophysical Prospecting, 2019, 54(2):288-301.
[13]
周晓华, 林君, 陈祖斌, 等.改进的神经网络反演微动面波频散曲线[J]. 吉林大学学报(地球科学版), 2011, 41(3):900-906.ZHOU Xiaohua, LIN Jun, CHEN Zubin, et al.Iterative inversion of microtremor surface wave dispersion curves by improved neural network[J]. Journal of Jilin University(Earth Science Edition), 2011, 41(3):900-906.
[14]
凡友华, 刘家琦, 肖柏勋.计算瑞利波频散曲线的快速矢量传递算法[J]. 湖南大学学报(自然科学版), 2002, 29(5):25-30.FAN Youhua, LIU Jiaqi, XIAO Boxun.Fast vector-transfer algorithm for computation of Rayleigh wave dispersion curves[J]. Journal of Hunan University(Natural Sciences), 2002, 29(5):25-30.
[15]
凡友华, 陈晓非, 刘雪峰, 等.Rayleigh波的频散方程高频近似分解和多模式激发数目[J]. 地球物理学报, 2007, 50(1):233-239.FAN Youhua, CHEN Xiaofei, LIU Xuefeng, et al.Approximate decomposition of the dispersion equation at high frequencies and the number of multimodes for Rayleigh waves[J]. Chinese Journal of Geophysics, 2007, 50(1):233-239.
[16]
尹晓菲, 胥鸿睿, 郝晓菡, 等.水平层状模型中多模式瑞雷波和拉夫波相速度频散曲线的灵敏度分析[J]. 石油地球物理勘探, 2020, 55(1):136-146.YIN Xiaofei, XU Hongrui, HAO Xiaohan, et al. Sensitivity analysis of multi-mode Rayleigh and Love wave phase-velocity dispersion curves in horizontal layered models[J]. Oil Geophysical Prospecting, 2020, 55(1):136-146.
[17]
张碧星, 肖柏勋, 杨文杰, 等.瑞利波勘探中"之"形频散曲线的形成机理及反演研究[J]. 地球物理学报, 2000, 43(4):557-567.ZHANG Bixing, XIAO Boxun, YANG Wenjie, et al.Mechanism of zigzag dispersion curves in Rayleigh exploration and its inversion study[J]. Chinese Journal of Geophysics, 2000, 43(4):557-567.
[18]
Dal Moro G, Pipan M, Gabrielli P.Rayleigh wave dispersion curve inversion via genetic algorithms and marginal posterior probability density estimation[J]. Journal of Applied Geophysics, 2007, 61(1):39-55.
