Inversion of Oceanic Parameters Represented by CTD Utilizing Seismic Multi-Attributes Based on Convolutional Neural Network

In Recent years,seismic data have been widely used in seismic oceanography for the inversion of oceanic parameters represented by conductivity temperature depth(CTD).Using this technique,researchers can identify the water structure with high horizontal resolution,which compensates for the deficiencies of CTD data.However,conventional inversion methods are modeldriven,such as constrained sparse spike inversion(CSSI)and full waveform inversion(FWI),and typically require prior deterministic mapping operators.In this paper,we propose a novel inversion method based on a convolutional neural network(CNN),which is purely data-driven.To solve the problem of multiple solutions,we use stepwise regression to select the optimal attributes and their combination and take two-dimensional images of the selected attributes as input data.To prevent vanishing gradients,we use the rectified linear unit(ReLU)function as the activation function of the hidden layer.Moreover,the Adam and mini-batch algorithms are combined to improve stability and efficiency.The inversion results of field data indicate that the proposed method is a robust tool for accurately predicting oceanic parameters.

Modeling of the Shale Volume in the Hendijan Oil Field Using Seismic Attributes and Artificial Neural Networks

Petrophysical properties have played an important and definitive role in the study of oil and gas reservoirs,necessitating that diverse kinds of information are used to infer these properties.In this study,the seismic data related to the Hendijan oil field were utilised,along with the available logs of 7 wells of this field,in order to use the extracted relationships between seismic attributes and the values of the shale volume in the wells to estimate the shale volume in wells intervals.After the overall survey of data,a seismic line was selected and seismic inversion methods(model-based,band limited and sparse spike inversion)were applied to it.Amongst all of these techniques,the model-based method presented the better results.By using seismic attributes and artificial neural networks,the shale volume was then estimated using three types of neural networks,namely the probabilistic neural network(PNN),multi-layer feed-forward network(MLFN)and radial basic function network(RBFN).

Seismic impedance inversion based on cycle-consistent generative adversarial network

Deep learning has achieved great success in a variety of research fields and industrial applications.However,when applied to seismic inversion,the shortage of labeled data severely influences the performance of deep learning-based methods.In order to tackle this problem,we propose a novel seismic impedance inversion method based on a cycle-consistent generative adversarial network(Cycle-GAN).The proposed Cycle-GAN model includes two generative subnets and two discriminative subnets.Three kinds of loss,including cycle-consistent loss,adversarial loss,and estimation loss,are adopted to guide the training process.Benefit from the proposed structure,the information contained in unlabeled data can be extracted,and adversarial learning further guarantees that the prediction results share similar distributions with the real data.Moreover,a neural network visualization method is adopted to show that the proposed CNN model can learn more distinguishable features than the conventional CNN model.The robustness experiments on synthetic data sets show that the proposed method can achieve better performances than other methods in most cases.And the blind-well experiments on real seismic profiles show that the predicted impedance curve of the proposed method maintains a better correlation with the true impedance curve.

Seismic-inversion method for nonlinear mapping multilevel well–seismic matching based on bidirectional long short-term memory networks

In this paper,the recurrent neural network structure of a bidirectional long shortterm memory network(Bi-LSTM)with special memory cells that store information is used to characterize the deep features of the variation pattern between logging and seismic data.A mapping relationship model between high-frequency logging data and low-frequency seismic data is established via nonlinear mapping.The seismic waveform is infinitely approximated using the logging curve in the low-frequency band to obtain a nonlinear mapping model of this scale,which then stepwise approach the logging curve in the high-frequency band.Finally,a seismic-inversion method of nonlinear mapping multilevel well–seismic matching based on the Bi-LSTM network is developed.The characteristic of this method is that by applying the multilevel well–seismic matching process,the seismic data are stepwise matched to the scale range that is consistent with the logging curve.Further,the matching operator at each level can be stably obtained to effectively overcome the problems that occur in the well–seismic matching process,such as the inconsistency in the scale of two types of data,accuracy in extracting the seismic wavelet of the well-side seismic traces,and multiplicity of solutions.Model test and practical application demonstrate that this method improves the vertical resolution of inversion results,and at the same time,the boundary and the lateral characteristics of the sand body are well maintained to improve the accuracy of thin-layer sand body prediction and achieve an improved practical application effect.

An evaluation of deep thin coal seams and water-bearing/resisting layers in the quaternary system using seismic inversion

Non-liner wave equation inversion,wavelet analysis and artificial neural networks were used to obtain stratum parameters and the distribution of thin coal seams.The lithology of the water-bearing/resisting layer in the Quaternary system was also predicted.The implementation process included calculating the well log parameters,stratum contrasting the seismic data and the well logs,and extracting,studying and predicting seismic attributes.Seismic inversion parameters,including the layer velocity and wave impedance,were calculated and effectively used for prediction and analysis.Prior knowledge and seismic interpretation were used to remedy a dearth of seismic data during the inversion procedure.This enhanced the stability of the inversion method.Non-linear seismic inversion and artificial neural networks were used to interpret coal seismic lithology and to study the water-bearing/resisting layer in the Quaternary system.Interpretation of the 1～2 m thin coal seams,and also of the water-bearing/resisting layer in the Quaternary system,is provided.The upper mining limit can be lifted from 60 m to 45 m.The predictions show that this method can provide reliable data useful for thin coal seam exploitation and for lifting the upper mining limit,which is one of the principles of green mining.

Application of Seismic Inversion Using Logging Data as Constraints in Coalfield

Seismic inversion and basic theory are briefly presented and the main idea of this method is introduced. Both non-linear wave equation inversion technique and Complete Utilization of Samples Information (CUSI) neural network analysis are used in lithological interpretation in Jibei coal field. The prediction results indicate that this method can provide reliable data for thin coal exploitation and promising area evaluation.

Stratal Carbonate Content Inversion Using Seismic Data and Its Applications to the Northern South China Sea

On the basis of the relationship between the carbonate content and the stratal velocity and density, an exercise has been attempted using an artificial neural network on high-resolution seismic data for inversion of carbonate content with limited well measarements as a control. The method was applied to the slope area of the northern South China Sea near ODP Sites 1146 and 1148, and the results are satisfaetory. Before inversion calculation, a stepwise regression method was applied to obtain six properties related most closely to the carbonate content variations among the various properties on the seismic profiles across or near the wells. These include the average frequency, the integrated absolute amplitude, the dominant frequency, the reflection time, the derivative instantaneous amplitude, and the instantaneous frequency. The results, with carbonate content errors of mostly ±5 % relative to those measured from sediment samples, show a relatively accurate picture of carbonate distribution along the slope profile. This method pioneers a new quantitative model to acquire carbonate content variations directly from high-resolution seismic data. It will provide a new approach toward obtaining substitutive high-resolution sediment data for earth system studies related to basin evolution, especially in discussing the coupling between regional sedimentation and climate change.

Model-data-driven seismic inversion method based on small sample data

As sandstone layers in thin interbedded section are difficult to identify,conventional model-driven seismic inversion and data-driven seismic prediction methods have low precision in predicting them.To solve this problem,a model-data-driven seismic AVO(amplitude variation with offset)inversion method based on a space-variant objective function has been worked out.In this method,zero delay cross-correlation function and F norm are used to establish objective function.Based on inverse distance weighting theory,change of the objective function is controlled according to the location of the target CDP(common depth point),to change the constraint weights of training samples,initial low-frequency models,and seismic data on the inversion.Hence,the proposed method can get high resolution and high-accuracy velocity and density from inversion of small sample data,and is suitable for identifying thin interbedded sand bodies.Tests with thin interbedded geological models show that the proposed method has high inversion accuracy and resolution for small sample data,and can identify sandstone and mudstone layers of about one-30th of the dominant wavelength thick.Tests on the field data of Lishui sag show that the inversion results of the proposed method have small relative error with well-log data,and can identify thin interbedded sandstone layers of about one-15th of the dominant wavelength thick with small sample data.

High resolution pre-stack seismic inversion using few-shot learning

We propose to use a Few-Shot Learning(FSL)method for the pre-stack seismic inversion problem in obtaining a high resolution reservoir model from recorded seismic data.Recently,artificial neural network(ANN)demonstrates great advantages for seismic inversion because of its powerful feature extraction and parameter learning ability.Hence,ANN method could provide a high resolution inversion result that are critical for reservoir characterization.However,the ANN approach requires plenty of labeled samples for training in order to obtain a satisfactory result.For the common problem of scarce samples in the ANN seismic inversion,we create a novel pre-stack seismic inversion method that takes advantage of the FSL.The results of conventional inversion are used as the auxiliary dataset for ANN based on FSL,while the well log is regarded the scarce training dataset.According to the characteristics of seismic inversion(large amount and high dimensional),we construct an arch network(A-Net)architecture to implement this method.An example shows that this method can improve the accuracy and resolution of inversion results.

Forward prediction for tunnel geology and classification of surrounding rock based on seismic wave velocity layered tomography

Excavation under complex geological conditions requires effective and accurate geological forward-prospecting to detect the unfavorable geological structure and estimate the classification of surround-ing rock in front of the tunnel face.In this work,a forward-prediction method for tunnel geology and classification of surrounding rock is developed based on seismic wave velocity layered tomography.In particular,for the problem of strong multi-solution of wave velocity inversion caused by few ray paths in the narrow space of the tunnel,a layered inversion based on regularization is proposed.By reducing the inversion area of each iteration step and applying straight-line interface assumption,the convergence and accuracy of wave velocity inversion are effectively improved.Furthermore,a surrounding rock classification network based on autoencoder is constructed.The mapping relationship between wave velocity and classification of surrounding rock is established with density,Poisson’s ratio and elastic modulus as links.Two numerical examples with geological conditions similar to that in the field tunnel and a field case study in an urban subway tunnel verify the potential of the proposed method for practical application.

致密碎屑岩储层地震反演技术方案及应用

致密碎屑岩储层与围岩地震波阻抗差异微弱,应用常规储层地震反演方法的有效性较差,精度较低。针对致密碎屑岩储层的地质、地球物理特点,本文提出了将神经网络和地质统计学结合起来的致密碎屑岩储层地震反演技术方案。在此方案中,通过神经网络地震反演获得地质涵义较为明确的但垂向精度较低的反演结果,以此结果为约束,以测井数据作为条件数据(硬数据)进行储层参数地质统计学随机反演/模拟,进而得到较为精细的、同时横向分布较为符合地质规律的储层参数反演成果。通过在D气田致密碎屑岩储层地震反演中的应用,提高了储层预测精度,达到了良好的应用效果。

砂砾岩储集层的地震反演方法

T719井区位于东营凹陷宁海断裂构造带的T94断层和胜北断层交汇区下降盘 ,沙四段—沙三段沉积期的古地貌沟梁相间 ,发育了夹于烃源岩内的规模较大的深水浊积扇体 ,主要储集岩为含砾砂岩和砾岩 ,其厚度差异较大 ,横向变化非常快。为研究储集层的变化细节并精细预测其分布 ,采用经高分辨率处理的纯波带地震数据 ,以测井资料为约束条件 ,用神经网络技术建立精细波阻抗初始模型 ,再利用拟线性方法细微调整 ,在最终反演出的具有较宽频带的绝对波阻抗剖面上 ,砂砾岩扇体与波阻抗的对应关系良好 ,其分布范围与实际钻探结果吻合。图 3参

基于BP神经网络的波阻抗反演及应用

人工神经网络是近期发展最快的人工智能领域研究成果之一.本文在介绍BP神经网络的有关原理的基础上,提出一种基于BP神经网络模型的波阻抗反演方法,该方法克服了常规基于模型的波阻抗反演方法严重依赖于初始模型的选择和易陷入局部最优等局限性.利用该方法对实际地震剖面进行了波阻抗参数反演处理,结果表明人工神经网络方法在波阻抗反演中的应用是可行的并且是有效的.

GPS掩星技术和电离层反演

系统地介绍了GPS无线电掩星技术的发展现状、系统组成和一些关键技术;论述了利用GPS掩星数据进行电离层反演的理论、模型及相应的计算流程,并结合GPS/MET和CHAMP卫星的实测资料,计算了电子密度剖面,与用其他方法所得结果的对比表明,GPS掩星电离层观测具有精度高、覆盖范围大等特点;最后讨论了GPS掩星技术应用于地震前兆监测的机理和前景。

叠后地震反演方法联合应用研究

叠后地震反演方法很多,不同的反演方法具有不同的优缺点和用途。在叠后反演过程中,通常只选用一种方法进行一次反演,反演效果往往难以满足储层预测的精度要求。通过采用递推反演、约束稀疏脉冲反演和神经网络反演3种方法联合反演,不断提高反演结果的分辨率,增加信息量。反演获得高精度的波阻抗、储层地球物理特征参数等多种数据体,可为储层预测提供高精度的资料。实际应用表明,该反演技术流程是可行、有效和实用的。

地震和测井联合反演储层波阻抗技术

通过地震和测井联合反演,可以获得高分辨率的井间地层波阻抗分布的信息。综合应用波动方程反演和神经网络分析来反演地层波阻抗参数,其过程分3步:第一步,应用先验地质知识,对地震数据和测井曲线进行地质解释,并对测井曲线进行对层和标定,然后求取相应的层速度的低频信息,旨在搞清井间地层结构状况,为非线性反演提供地层产状的先验信息;第二步,应用非线性波动方程反演,在层速度界面及井中物性参数约束下,从地震数据中反演高分辨率的反射系数及波阻抗参数;第三步,应用CUSI神经网络分析方法,以高分辨率的反射系数及波阻抗等参数作为约束,以沿层求取的地震特征作为输入,以井中反演的波阻抗参数为期望输出,对非线性波动方程反演出的波阻抗参数进行非线性标定,得出井间的地层绝对波阻抗物性参数。

支持向量机在重震联合反演中的应用研究

本文对利用支持向量机进行重震联合反演问题做了深入研究,特别是对支持向量机用于重震联合反演时的重力资料的处理、参数选取、特征量的提取等具体实现问题进行了讨论,最后用所设计的支持向量机重震联合反演模型对东营北部区域结晶基底岩做了预测反演,取得了满意的效果.

基于神经网络的随机地震反演方法

针对随机地震反演中存在的两个主要问题,随机实现含有噪声和难以从大量随机实现中挖掘有效信息,提出了一种基于神经网络的随机地震反演方法.通过对多组随机实现及其正演地震数据的计算,构建了基于序贯高斯模拟的训练集.这也为应用神经网络求解地球物理反问题,提供了一种有效建立训练集的方法.较之传统的神经网络反演,这种训练集不仅保证了学习样本具有多样性,同时还引入了空间相关性.数值模拟结果表明,该方法只需要通过单层前馈神经网络,就可以比较有效的解决一个500个阻抗参数的反演问题.

地震反演与属性耦合检测薄层含气砂岩

地震资料气层检测的常用方法是沿目的层拾取选定时窗内的地震属性进行属性分析,但该方法用于薄层砂岩储层的油气检测则比较困难。为此,本文利用非线性随机反演方法精细刻画砂体的空间展布,得到储层顶、底层位信息;再沿层提取各种地震属性,时窗随储层厚薄变化;分别应用基于RS理论的属性优化方法和SOM神经网络模式识别方法进行气层检测。将该方法应用于川西拗陷洛带气田,符合率达85%以上。

用模拟退火神经网络技术进行波阻抗反演

利用基于模拟退火算法的神经网络技术进行测井约束的波阻抗反演，可根据数据本身之间的内在联系建立一个自适应非线性认知系统，只要在输入端输入特征数据，便能在输出端得到期望输出值，而不必关心系统本身的内部机理。在反演前，从测井资料中整理出地层波阻抗参数，用神经网络建立起地震波特征和地层波阻抗参数的映射关系，然后再利用这种映射关系进行外推，得到其它地震道所对应的波阻抗参数。在训练过程中，引入了模拟退火算法，使网络能有效地避开局部极小，这样可以提高收敛速度和拟合精度。