全张量重力梯度数据的综合分析与处理解释

发布时间：2018-06-24 14:12

本文选题：全张量重力梯度 + 加强重力　；参考：《吉林大学》2015年博士论文

【摘要】：地球重力场是地球的固有物理特征，可以反映地球内部的物质分布、运动和变化的规律。重力测量是对重力场变化规律的直接反映。随着测量手段的多样化（地面测量、海洋测量、航空测量和卫星测量）和测量数据精度的提高，人们不仅可以直接测量地球重力场，而且可以对重力梯度场进行测量。重力梯度张量数据相对于传统的重力测量数据有着更高的频率信息，能更加准确、细致的研究地球浅部构造、矿产资源分布等信息。为了更加准确的处理和解释全张量重力梯度数据，需要从全张量重力梯度仪器设计原理出发，分析仪器产生噪声的原因，进而更好的从测量数据中提取反应地下异常体的真实信号进行解释。首先，本文利用理论模型对比分析了重力梯度测量相对于重力测量的优势，为文章的选题提供依据。重力梯度张量数据的获取，可以通过实际测量或计算得到。然而，计算得到的重力梯度张量数据最多能包含与测量的重力数据相同的信息成分，而且在计算过程中还会造成信息缺失，并不会增加信息成分。因此，重力梯度张量数据的获取通常需要通过重力梯度仪器测量得到。通过对比分析，重力梯度异常能反映相对波长短的高频信号，重力异常数据能反映相对波长长的低频信号。因此，联合梯度张量数据的高频信号成分和重力异常的低频信号成分可以得到加强的重力异常数据，该数据能同时保留重力异常数据及其梯度张量数据的所有信息，拓宽重力异常数据的频宽范围。这一过程主要通过余弦平方滤波完成。目前，国内外多家研究机构正在研究多种不同类型的重力梯度仪，但是，投入生产使用的只有基于旋转加速度计的重力梯度仪。因此，本文针对用于航空移动平台高精度全张量重力梯度测量系统，深入研究了全张量梯度仪12个加速度计按3个不同旋转轴圆盘形成差分组合的结构。在确认该结构具有有效抑制运动共模加速度减少外界环境干扰实现高精度探测优点的同时，着重分析了重力梯度仪测量误差的来源和影响。研究表明，主要影响包括仪器固有随机噪声和外界确定性噪声。为定量描述影响程度，笔者推导了在航空动态环境下的测量方程，并分离出加速度计的性能不匹配、平台不稳定、圆盘转速不稳定3个主要固有因素，从时间域和频率域角度定量分析固有影响因素的噪声水平。试验分析表明，通过利用Simulink仿真系统可以获得固有因素产生的噪声水平，并提出抑制方案。针对搭载环境测量误差，笔者还分析了实测飞行中姿态和质量的改变对重力梯度测量值造成的环境影响，提出了基于点质量源的自身梯度校正方法。经固有噪声和自身梯度校正后的各个梯度数据中还存在大量随机噪声，但是其噪声水平各不相同。本文利用奇偶测线网格化方法对各个梯度张量数据的噪声进行了定量估计，为后续噪声去除提供依据。全张量重力梯度张量数据相对于重力数据含有更高频的信号成分，能更好的描绘小的异常特征。然而，全张量重力梯度仪测量值的噪声成分也为高频。因此，从高频信号成分中分离出噪声将是处理重力梯度张量数据的一个挑战。通过联合全张量重力梯度仪测量的多个信息成分进行噪声处理能更好的压制随机干扰。本文利用重力梯度信号满足的拉普拉斯方程约束条件，推导了重力梯度张量的笛卡尔方程和引力位的级数解表达式，然后利用级数解去拟合测量的重力梯度张量值，拟合的部分认为是真实的梯度信号，未拟合的部分认为是噪声。在拟合过程中引入数据噪声加权矩阵和数据能量加权矩阵进行最优线性反演求解拉普拉斯方程级数解的系数，然后利用求得的系数进行正演计算得到真实的梯度信号。全张量重力梯度数据经噪声滤波后，只包含实际地下重力信息，能更加准确的进行数据的解释。重力梯度数据的解释工作通常需要从数据中获取有关场源的水平位置和深度范围。水平位置的确定通常利用边界识别方法；深度范围参数的确定需要深度计算方法来完成。边界识别方法在重力梯度数据解释中占有重要角色，它能准确且快速地确定地质体边界位置、构造水平位置而被广泛关注。一些传统的方法有的不能同时显示不同埋藏深度地质体的边界信息，有的在深部地质体边界位置确定中误差较大，且对细节信息的提取能力不足。而且，已有的方法基本上都是针对重力异常数据，专门针对重力梯度张量数据的方法还比较少。本文针对全张量重力梯度数据信息量大、信号频率高，能更好的描述小的异常特征等特点，提出了改进的水平解析信号方法、加强水平方向总水平导数方法和改进的结构张量算法进行全张量重力梯度数据的边界解释。在解释过程中，本文针对各个方法进行相应的归一化处理，使其能均衡不同埋藏深度的异常体的边界结果的振幅强度。通过模型试验和实际测量的全张量重力梯度数据验证了这些方法的可靠性、实用性和抗噪能力，并与一些传统的边界识别方法进行对比，证明了这些方法的优点。重力异常及重力梯度异常场源深度计算的方法经历了漫长的发展历史，形成了针对不同的场源类型和地质研究对象的不同深度计算方法。快速成像方法是近几年的一个发展热点，它能快速得到地下异常体分布状况，避免传统方法耗时长，内存消耗大等缺点。本文利用极大值深度计算方法DEXP进行地下异常体的埋藏深度成像，从而进行异常体的深度估计。然而，，传统的DEXP方法需要事先指定异常体的构造指数，通常是根据异常形态对其进行假设，但这会对成像结果带来误差。因此，本文又利用不同阶的垂向导数的比值进行DEXP变换，有效的去除了构造指数的影响。该方法能在不知道构造指数的情况下对异常体进行深度成像，从而得到地质体的深度。利用估计的深度值在尺度函数上的对应值可以对地质体的构造指数进行估计。利用该方法对模型数据和实际测量的文顿岩丘的重力梯度数据进行分析，取得了准确的深度和构造指数估计。
[Abstract]:The earth's gravity field is an inherent physical feature of the earth, which can reflect the distribution of material, movement and change of the earth's interior. Gravity measurement is a direct reflection of the law of the variation of the gravity field. With the diversification of the measuring means (ground measurement, oceanographic survey, aeronautical measurement and satellite measurement) and the accuracy of measurement data, people can not only be able to improve the accuracy of the measurement data. The gravity gradient field can be measured directly and the gravity gradient field can be measured. The gravity gradient tensor data has higher frequency information relative to the traditional gravimetric data. It can be more accurate and detailed to study the shallow structure of the earth and the distribution of mineral resources. In order to more accurately deal with and explain the full tensor gravity gradient. According to the design principle of the full tensor gravity gradient instrument, the cause of the noise generated by the instrument is analyzed, and then the real signal of the underground abnormal body is extracted from the measured data.
First, in this paper, the advantages of gravity gradient measurement relative to gravity measurement are compared and analyzed by the theoretical model. The acquisition of the gravity gradient tensor data can be obtained by actual measurement or calculation. However, the calculated gravity gradient tensor data can contain most of the same data as the measured gravity data. Therefore, the acquisition of gravity gradient tensor data is usually measured by gravity gradient instruments. By contrast, gravity gradient anomalies can reflect high frequency signals with short relative wavelengths, and gravity anomaly data can reflect the length of relative wavelengths. Therefore, the high frequency signal components of the combined gradient tensor data and the low frequency signal components of the gravity anomaly can obtain enhanced gravity anomaly data. This data can simultaneously retain all the information of the gravity anomaly data and its gradient tensor data, widening the bandwidth of the gravity anomaly data. This process is mainly through the cosine square. The filter is completed.
At present, many research institutions at home and abroad are studying a variety of different types of gravity gradiometer. However, only the gravity gradiometer based on the rotation accelerometer is used in production. Therefore, this paper has studied 12 accelerometers by full tensor gradiometer for the high precision full tensor gravity gradient measurement system used in the aviation Mobile platform. 3 disks of different rotating axes form a structure of difference combinations. The source and influence of the measurement error of gravity gradiometer are emphatically analyzed in the confirmation that the structure has the advantages of effective suppression of the movement common mode acceleration and the reduction of external environment interference. The main influence of the structure is that the main influence includes the inherent random noise and the outside accuracy of the instrument. Qualitative noise. In order to quantitatively describe the degree of influence, the author derives the measurement equation under the air dynamic environment, and separates the 3 main inherent factors of the accelerometer's performance mismatch, the instability of the platform and the instability of the disc speed, and the analysis of the noise level which has the influence factors from the time domain and the frequency domain. The noise level generated by inherent factors can be obtained by using the Simulink simulation system, and the suppression scheme is proposed. In view of the environmental measurement error carrying the environment, the author also analyzes the environmental influence caused by the change of attitude and mass to the measured value of gravity gradient in the measured flight, and proposes a self gradient correction method based on the point mass source. There are a lot of random noises in each gradient data after sound and self gradient correction, but their noise levels are different. This paper quantificationally estimates the noise of each gradient tensor data by using the odd even grid method to provide the basis for the subsequent noise removal.
The full tensor gravity gradient tensor data contains a more high-frequency signal component relative to the gravity data, which can better describe the small anomaly characteristics. However, the noise component of the total tensor gravity gradiometer is also high frequency. Therefore, the separation of noise from the high frequency signal component will be a challenge to deal with the gravity gradient tensor data. The multiple information components measured by the full tensor gravity gradiometer can better suppress random interference. This paper derives the expression of the Cartesian equation of gravity gradient tensor and the series of gravitational potential, and then uses the series solution to fit the measured gravity with the constraints of the Laplasse equation satisfied by the gravity gradient signal. The value of the gradient tensor is considered to be a real gradient signal, and the part of the non fitting is considered to be noise. In the fitting process, the data noise weighting matrix and the data energy weighting matrix are introduced to the optimal linear inversion to solve the coefficients of the series solution of the Laplasse equation, and then the calculated coefficients are calculated to be true. Gradient signal. After noise filtering, full tensor gravity gradient data only contains actual underground gravity information, it can be more accurate to explain the data. The interpretation work of gravity gradient data usually needs to obtain the horizontal and depth range of the related field sources from the data. The determination of range parameters requires deep calculation.
The boundary recognition method plays an important role in the interpretation of gravity gradient data. It can accurately and quickly determine the location of the geological body boundary and construct the horizontal position. Some traditional methods can not display the boundary information of different buried depth geological bodies at the same time, and there are some errors in the determination of the boundary position of the deep geological body. In addition, the existing methods are basically aimed at gravity anomaly data, and the methods specially aimed at gravity gradient tensor data are relatively few. In this paper, the characteristics of the full tensor gravity gradient data are large, the signal frequency is high, and the small abnormal characteristics can be described better. The improved water is proposed. The horizontal analytic signal method, the horizontal directional total horizontal derivative method and the improved structural tensor algorithm are used to explain the boundary of the full tensor gravity gradient data. In the process of interpretation, this paper deals with the corresponding normalization of each method so that it can balance the amplitude intensity of the boundary result of the abnormal body with different buried depth. The full tensor gravity gradient data of the model test and the actual measurement verify the reliability, practicability and anti noise ability of these methods, and compare with some traditional method of boundary recognition, which prove the advantages of these methods.
The method of calculating gravity anomaly and gravity gradient anomaly source depth has experienced a long history, forming different depth calculation methods for different field source types and geological research objects. Fast imaging method is a hot spot of development in recent years. It can quickly get the distribution of ground anomaly bodies and avoid the time-consuming of traditional methods. This paper uses the maximum depth calculation method DEXP to carry out the buried depth imaging of the underground abnormal body by using the maximum value depth calculation method, so as to estimate the depth of the abnormal body. However, the traditional DEXP method needs to specify the structure index of the abnormal body in advance, usually based on the abnormal form, but this will bring the imaging result. Therefore, this paper makes use of the ratio of the vertical derivative of different order to carry out DEXP transformation, effectively removing the influence of the tectonic index. This method can make the depth imaging of the abnormal body without knowing the structure index, so that the depth of the geological body can be obtained. The structure index of the body is estimated. The method is used to analyze the gravity gradient data of the model data and the actual measured venturi mound, and the accurate depth and structural index are estimated.
【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：P631.14

【参考文献】