A nonlinear microseismic source location method based on Simplex method and its residual analysis

Source location is one of the most valuable features of the microseismic technique due to its ability to delineate the unstable areas. In this paper, the precise formulas of the station residual and event residual are derived for the L1 norm statistical standard and the L2 norm statistical standard based on the residual analysis. Then, the error space for microseismic source location is proposed and analyzed. Based on the above research, a nonlinear microseismic source location method using the Simplex method is developed. This new method can search the microseismic source directly in the error space through four deformations of the simplex figures, and it is able to make use of both P-wave and S-wave velocities. Finally, the performance of the Simplex microseismic source location method is tested and verified by laboratory experiments. Test results show that the Simplex microseismic source location method can improve the accuracy and stability of the source location greatly when P-wave and S-wave velocities are involved simultaneously and correctly. The results also demonstrate that the L1 norm statistical standard always provides more accurate and reliable solutions than the L2 norm statistical standard when there are some major but isolated errors in the input data. However, none of the optimization methods are able to function when the errors in the input data are systematic and extreme, which indicates that an early detection and correction of these errors is of primary importance for microseismic source location.     

微地震震源定位方法综述

微地震震源定位方法是微地震监测领域的一项核心技术,而考量微地震技术应用效果好坏的准则在于震源定位方法的精确程度。针对非常规油气开发过程中微地震震源定位方法的应用,本文对微地震震源定位方法进行阐述。其中：几何作图法具有稳健、效率高的优势,但震源位置较深时定位精度较低;线性定位法无需速度模型精度,但对初至拾取的精度有较大影响;非线性定位法对初至拾取较为敏感,对速度模型的精度要求较高,但计算量较小;混合优化定位法在一定程度上提高了定位的精度和效率,但在低信噪比、速度模型精度较低时优势不明显;基于波形偏移的定位方法无需考虑初至拾取的精度,但计算量较大;基于神经网络的定位方法采用训练网络进行训练,定位精度高,误差小。同时,本文还介绍了多方法多参数信息融合技术在油气藏微震震源定位中的应用。     

基于稳健模拟退火-单纯形混合算法的微震定位研究

为提高矿井微震的定位精度,通过模拟试验对比分析常规定位及新方法在定位求解中的适用范围及优缺点,研究反演未知变量数目、检波器的密度、波速等对各自定位方法的影响。基于单纯形法的优点,并结合模拟退火法的全局收敛性,提出采用稳健的模拟退火-单纯形法进行微震定位。研究发现,该方法不必直接搜索发震时刻,只需通过稳健方法预先估算,利用较少的检波器就可进行微震定位,且基本不受初始波速影响,从而大大减少搜索步数,有效地提高定位精度及收敛速度。以柿竹园矿人工爆破定位测试试验为例,重点分析比较模拟退火-单纯形法、稳健模拟退火-单纯形法及其他定位方法的空间绝对距离及目标函数误差,认为检波器数量较多且可获得理想的P波信息时,应用非线性最小二乘法能求出可信的结果;其余定位方法的三维定位误差及目标函数误差相差不大,但稳健模拟退火-单纯形法目标函数误差最小,稳定性最高,可在矿井微震定位中推广使用。     

基于多道信噪能量统计的微地震有效信号识别方法

在微地震监测中有效微地震信号的识别是关键步骤。常规的微地震有效信号识别方法基于单道长短时窗能量比,受噪音影响较大,对于信噪比较低的事件无法准确识别。利用多道记录数据之间的信号相关性好而信号与噪音相关性差的特点,通过记录道自相关求取总能量和相邻记录互相关求取信号能量,以有效信号能量所占比即含信能量比作为门槛值进行有效信号识别,信号识别过程中利用多道互相关对噪音进行了有效压制,提高了微地震信号识别的压噪能力,通过模型数据和实际资料处理验证了方法的有效性。     

利用P波S波相结合的逆时偏移方法监测微地震群

微地震定位是油气资源开采中评估水力压裂质量和评估矿山地震和水库地震安全的重要手段。但是,采用地震震相 到时定位的方法很难对具有相近到时的微地震事件群精确定位地震。本文尝试用P波和S波结合的逆时偏移方法对具有相 近到时的微地震定位。先用绝对振幅记录做逆时偏移,获得P波和S波以及P波与S波结合的综合波场,再采用波前在震源 位置成像前后的汇聚到散射的转换过程作为成像条件,提取震源位置和发震时刻。用数值模拟生成具有相近到时的微地震 群,然后用该方法定位微地震。结果表明该方法可以有效提取各个震源位置和发震时刻。     

基于波形参数的微震P波到时拾取值质量控制方法

微震事件中常常包含一些异常信号、强噪声干扰信号和弱信号,这些通道信号的P波到时自动拾取精度往往很低,甚至拾取错误。目前国内外微震监测系统进行自动定位时,并不对各个P波到时拾取值预先筛选,而需要技术人员手动剔除或修正部分无效P波到时,然后才能进行正确定位。为此,首先引入了一种Akaike信息准则（AIC）两步骤拾取算法,并基于某深埋隧道微震监测案例定量分析了P波到时拾取误差与震源定位精度之间的关系,从而引入了P波到时拾取值容许误差的概念。然后,利用AIC两步骤拾取算法拾取微震波形的P波到时,并计算各项波形参数,进行大量统计,深入研究了影响P波到时拾取精度的波形参数。以容许误差为基准,将P波到时拾取值分为有效（标签为1）和无效（标签为?1）两类,并以相应的波形参数为数据输入,采用支持向量机方法（SVM）训练数据,最终建立了P波到时拾取值质量控制模型。实际应用表明：P波到时拾取值质量控制方法能有效剔除错误拾取值,从而大幅度提高数据自动处理效率和震源定位精度。     

微地震震源定位中目标函数的研究与应用

针对常规微地震震源定位中走时目标函数的多解性问题,基于三分量微地震数据,建立了联合走时和三分量微震数据方位信息的矢量目标函数,把矢量目标函数运用于Geiger定位方法中,得到了改进的Geiger定位法。模型数据测试表明:该方法对震源的方位信息敏感,定位精度更高。      

Practical application of an innovative microseismic source location procedure

Summary A detailed case study conducted at a Canadian gold mine is presented to demonstrate the practical usefulness of the source location data analysis theory developed by Ge and Kaiser (1990). The basic concept of the theory and the interpretation of the physical status of wave arrival picks is explained through a step-by-step analysis of several microseismic events. The effectiveness of the theory is illustrated by statistical data analysis. It is demonstrated that the theory enhances MP250 systems by providing an automatic microseismic source location technique.     

Sediment type distribution in reservoirs: sediment source versus morphometry

 Bottom sediments were collected from seven Texas reservoirs to assess the spatial distribution of sediment types in reservoirs. The sediment samples were analyzed for water content, organic content, and grain-size distributions. The reservoirs are characterized by distinct water content patterns (referred to in this paper as Types I, II, and III) that reflect the lithologic units comprising their drainage area. The water content, organic content, and percent mud decreases from Type I (shale, marl) through Type II (limestone-marl-sandstone) to Type III (granite-gneiss-schist-sandstone) reservoirs. Reported elemental concentrations in the reservoir sediments are consistent with concentrations in the dominant rocks within their drainage area. While water depth accounts for part of the spatial distribution, sediment source explains the overall distribution of sediment types. Coarse-grained source rocks, multiple source rocks, and multiple tributaries which input sediment at different points contribute to an inconsistent sediment type distribution. Depending on the sediment source, sand and gravel-size sediments are present in the deeper regions of some reservoirs. This disrupts the classical morphometry (water depth, slope) controlled sediment distribution pattern (coarser sediments in shallower regions and finer sediments in deeper regions) of natural lakes. Thus, the location of tributaries and the type of sediments they carry are the dominant factors that control the sediment type distribution in reservoirs. Received: 16 June 1998 · Accepted: 24 August 1998     

基于差分进化算法的微震定位

针对传统基于走时微震定位方法存在的定位精度依赖于走时精度问题,提出了改进定位目标函数的差分进化微震定位方法。基于传统微震P波走时定位方法,在定位目标函数中加入了误差加权系数,通过降低初至拾取精度低的微震信号道对定位误差的影响,减弱了初至拾取精度对微震定位的影响;利用差分进化算法,通过种群初始化、变异、交叉和选择等操作,求解改进的微震定位目标函数,实现微震定位;对模型数据和实测数据进行试处理,结果表明:改进的目标函数能明显降低误差较大分量对总误差的贡献;提出的改进定位目标函数的差分进化微震定位方法,在部分微震信号道走时拾取误差较大甚至错误的情况下,能很大程度减弱拾取误差对定位精度的影响,可以获得较好的定位结果。      

基于快速行进迎风线性插值的微震定位算法研究

准确的微震定位方法是微震监测技术进行岩体稳定性分析的基础,而初至波走时计算方法是决定微震定位精度重要因素之一。引入一种由快速行进法(fast marching methods, FMM)与线性插值射线追踪法(LTI ray tracing)改进而来的线性插值算法,实现了在复杂岩体中相对准确的微震初至走时计算,并在均一速度模型和Marmousi模型中得到了验证。在此基础上,采用常微分方程数值算法中的单步高精度Runge-Kutta法,得到了震源点和传感器之间的射线路径,并在分层和带空洞的岩体模型中验证了其合理性。选取最小二乘法为目标函数,通过寻找目标函数值最小的方式来获得最优节点实现微震定位。在所建立的加入地下硐室的Marmousi岩体模型中,对于预设震源点的定位结果误差分别为1.41、1.00、2.83、1.41、1.41m,验证了所提方法在复杂岩体速度模型中能够实现较准确的微震定位,为进一步提高工程岩体的微震定位精度提供了新思路。     

A Mathematical Calculation Model Using Biomarkers to Quantitatively Determine the Relative Source Proportion of Mixed Oils

It is difficult to identify the source（s） of mixed oils from multiple source rocks, and in particular the relative contribution of each source rock. Artificial mixing experiments using typical crude oils and ratios of different biomarkers show that the relative contribution changes are non-linear when two oils with different concentrations of biomarkers mix with each other. This may result in an incorrect conclusion if ratios of biomarkers and a simple binary linear equation are used to calculate the contribution proportion of each end-member to the mixed oil. The changes of biomarker ratios with the mixing proportion of end-member oils in the trinal mixing model are more complex than in the binary mixing model. When four or more oils mix, the contribution proportion of each end-member oil to the mixed oil cannot be calculated using biomarker ratios and a simple formula. Artificial mixing experiments on typical oils reveal that the absolute concentrations of biomarkers in the mixed oil cause a linear change with mixing proportion of each end-member. Mathematical inferences verify such linear changes. Some of the mathematical calculation methods using the absolute concentrations or ratios of biomarkers to quantitatively determine the proportion of each end-member in the mixed oils are deduced from the results of artificial experiments and by theoretical inference. Ratio of two biomarker compounds changes as a hyperbola with the mixing proportion in the binary mixing model, as a hyperboloid in the trinal mixing model, and as a hypersurface when mixing more than three end- members. The mixing proportion of each end-member can be quantitatively determined with these mathematical models, using the absolute concentrations and the ratios of biomarkers. The mathematical calculation model is more economical, convenient, accurate and reliable than conventional artificial mixing methods.     

微震震源定位新方法——混合差定位法

当微地震事件个数较多时,传统双差定位算法的数据存储量和计算量巨大,难以适应微震震源的定位。提出一种新的微震震源定位算法——混合差定位法。这种算法综合了双差定位算法和Geiger定位法的优点,不仅数据的存储量和计算量大大减小,且与双差定位相比,其定位结果不受初始震源丛的质心位置的影响。模型数据和实际微地震数据测试结果验证了该方法的有效性和实用性。      

爆破开挖诱发的地下交叉洞室微震特性及破裂机制分析

白鹤滩水电站右岸地下多洞室间相互施工扰动,使得母线洞等交叉洞室围岩的微震活动规律及破裂机制十分复杂。引进南非IMS微震监测系统,研究高地应力、多面临空卸荷应力环境下交叉洞室围岩的微震特性及破裂孕育机制。根据地下厂房区围岩岩性、错动带及断层分布,选取监测效果较佳的传感器并合理布置;通过定点敲击试验进行波速反演,分析震源定位的精度;以10#母线洞掉块案例为工程背景,分析爆破开挖诱发的交叉洞室岩体破裂微震事件与能量释放时空演化规律,并基于S波和P波的能量比（ES /EP）,归纳、总结了掉块发生过程中岩体破裂演化机制：大量张拉事件发生（爆破冲击诱发）→拉剪、压剪事件发生→张拉事件、剪切事件交替发生（集中应力和节理综合影响）→掉块发生（重力和节理综合作用）。研究结果有利于优化白鹤滩水电站右岸地下洞室群交叉部位的开挖方案,同时对类似工程的开挖和支护具有重要借鉴意义。     

用生物标志物定量计算混合原油油源的数学模型

多烃源层叠合盆地混源油油源的定量计算是十分困难的问题.典型原油的人工混合配比实验显示,当两个生物标志物浓度不同的原油相混合时,各类生物标志物的比值参数随端元原油混入量呈非线性变化,以生物标志物比值参数和简单的二元线性关系方程定量计算混源油的混合比例将导致错误的结论.三个或者三个以上原油相混合时,各类生物标志物比值参数的变化更加复杂,各端元原油的贡献更加难以判识.但是,人工混合模拟实验表明混合油中生物标志物绝对含量与端元油的混入量呈线性关系,数学推导证明了这种线性关系,由此推导出相应的数学计算模型,其中:二元混合时比值参数与混入量呈双曲线关系,三元混合呈双曲面关系,四元及其以上的多元混合呈多维曲面,可以矩阵的方式定量计算各端元油的比例.依据这些数学模型,应用生物标志物的绝对含量和(或)生物标志物比值参数均可以定量计算出混源原油中各类原油的贡献比例.数学模型比通常的人工模拟实验方法更加经济、方便、精确和可靠.     

含采空区层状岩体微震信号传播规律试验研究

微震信号在层状岩体中传播规律的研究对于实现震源准确定位具有重要的现实意义。在天然岩体中,微震信号在岩层中传播的速度,不仅受其内部因素即岩石自身物理性质的影响,还受外部因素如地质结构面、断层及采空区等条件的影响。通过室内试验的方法对微震信号在不同岩层及采空区处的传播规律进行研究。结果表明：波速随着传播距离的增加而逐渐衰减,且岩层密度越大衰减越慢;穿过的结构面数量越多,微震信号传播速度衰减比例越大;能量的衰减与波速的衰减一致。采空区断面横截面积越大,微震信号通过的用时越长;采空区周围岩石密度、弹性模量等物理性质的数值越大,微震信号通过时衰减越少。边际谱分析结果表明：断层对频率高的微震信号阻碍作用更强。     

鄂尔多斯盆地甘泉地区延长组长8段原油地球化学特征及来源

通过系统采集鄂尔多斯盆地东南部甘泉地区原油、砂岩和烃源岩样品,分析了延长组长8段原油地球化学特征,明确了原油成因和来源。研究表明:长8段原油为正常原油,具有低黏度、低初馏点、低凝点和低含硫量的特征;生物来源为混合型输入,形成于弱氧化-弱还原环境,是生油高峰阶段的产物。根据C₃₀重排藿烷和二环倍半萜类分布特征可将原油分为Ⅰ类和Ⅱ类原油,Ⅰ类原油主要分布于长8₁段,Ⅱ类原油主要分布于长8₂段。长9段和长8段烃源岩具低C₂₄TeT/C₂₆TT、高-异常高C₃₀重排藿烷含量、低8β(H)升补身烷和8β(H)补身烷含量的特征,长7段烃源岩与之相反。油源对比结果表明,Ⅰ类原油为长7烃源岩贡献,Ⅱ类原油为长8和长9段烃源岩贡献。甘泉地区主要发育长8₂油藏,其西部原油为长8和长9烃源岩贡献,而东部原油主要为长9烃源岩贡献,长7烃源岩次之。研究认识到长8和长9烃源岩的烃类贡献可能是促使长8₂段较长8₁段富集的重要因素,对下步勘探有指导意义。     

闽台地区地震速报工作回顾

本文指出了福建微震仪地震台网在地震速报和编目工作中存在的主要问题以及问题产生的可能原因;着重阐述了闽台两地区速度模型差异大是造成福建地震台网定位台湾地区震中误差的主原因;最后提出减少震中定位误差的改进办法。     

Interpretation of some magnetic bodies using neural networks inversion

Magnetic surveys have been used for mineral exploration where different data processing techniques were used to derive the parameters of causative targets. In this respect, the neural network (NN) technique was used to estimate the magnetic causative target parameters. Examples of NN inversion have been tested on synthetic examples where the NN was trained well using forward models of the vertical magnetic effect of a vertical sheet and a horizontal circular cylinder. Specifically, modular neural network (MNN) inversion has been used for the parameter estimation of the causative targets, where the sigmoid function was used as the activation function. The effect of random noise and the error estimation of the horizontal location have been analyzed. When NN is applied to real data, it estimates successfully the parameters of the causative targets such as burial depths, magnetic constants, and angle of polarization. Hilbert transform has been used to locate the source origin, which is important for the NN inversion. This approach has more advantages than the conventional data inversions in terms of its efficiency and flexibility. It also gives fast solutions. The MNN approach has been applied to the Kursk and Manjampalli anomalies, where the results were shown to be in good agreement with the other techniques published in the literature.