磁性目标小子域定位算法及相关技术研究
发布时间：2018-08-19 15:38
【摘要】：近年来,随着地磁传感器仪器的探测精度和探测速度的不断提高,人们可以获得大量的高精度地磁数据,这使得地磁探测在各种应用领域中有着更为广阔的应用前景。地磁探测中,确定目标物体的位置是一项首要任务,是进行后续工作的前提。由于铁磁性目标的存在,其自身的固有磁场和产生的感应磁场会叠加到地磁场上,从而导致空间地磁场分布的变化,进而在该空间中产生磁异常。人们通过对磁异常的分析和反演,可以有效的获得该目标物体的位置,达到定位的目的。因此研究如何利用磁异常进行磁性目标定位具有重要的学术和实际意义。高精度的磁测仪器(光泵磁力仪,超导量子干涉仪等)在国内外的地磁探测中得到广泛的应用。这些仪器具有较高的测量精度和稳定性,同时能够连续测量,为获得可靠的地磁数据提供了必要的保障。但是在利用磁异常进行目标定位过程中,仍然存在一些重要的技术难点,如:目标磁异常的提取,地磁图的空间插值技术,测量过程中噪声的抑制方法以及定位算法等。本论文针对上述问题展开研究,研究工作主要集中在以下几个方面:目标磁异常的提取是磁性目标定位中关键任务。地磁场可以分解为地磁正常场和异常场两部分,异常场是相对于正常场而言的,是由于磁性目标的存在而产生的。由于选取的地磁正常场不同,故计算出的异常场也不完全相同。为了获得准确的磁异常,需要构建合理的地磁正常场模型。地磁场模型分为全球模型与区域模型。相比于全球模型,区域模型更能够精细地描绘该区域内的磁场分布。准确构建区域的地磁正常场模型决定了目标磁异常提取的准确程度。本文分析了两种构建局部区域地磁场的模型:多项式模型和Spline模型,并对比了两种模型的优缺点,分析了两种模型存在的问题,给出了提高精度的应对措施和最佳模型选择方案。地磁图是描绘磁异常分布的手段和工具,同时地磁图数据是进行定位的关键。由于实际测量中很难能实现大规模的地磁测量条件而获得足够的测量数据,因此,需要通过空间插值方法来提高测量数据的密度,从而提高磁异常的精度。克里金插值法在表现属性的空间相关性的问题上有很高的逼近程度,因此,本文将克里金插值法做为地磁图绘中的插值方法。文中从数学原理出发,探析了该插值方法的数学原理。克里金插值方中的变异函数则能够反映局部范围和特定方向上特征的变化,因此,变异函数决定了插值精度。插值过程中的变异函数是实验变异函数和理论变异函数的拟合结果。本文提出了基于粒子群算法的变异函数拟合法,有效提高了变异函数的拟合精度,从而提高了地磁图绘制的准确程度。通过实验验证了基于粒子群算法拟合得到的变异函数在插值过程中的效果好,插值得到的地磁图的精度也很高。实际地磁测量过程中的噪声会对测量数据造成干扰,影响测量数据的精度。为了提高数据的质量,需要消除地磁信号中的噪声。地磁噪声的频带较宽,很难通过频率分离的方式将噪声分离处理。由于信号和噪声在小波域有不同的表现形式,小波去噪是通过它们的小波系数幅值随尺度变化的趋势不同,达到分离噪声的目的。本文针对小波阈值去噪中传统的阈值函数存在着一些缺点,提出了一种阈值可以动态调整的阈值函数,有效避免软阈值方法中对于绝对值较大的小波系数与其处理后的小波系数估计值相比总有固定偏差,且产生恒定衰减的不足。同时针对如何选择合适的阈值问题,给出了阈值选取的方法。仿真结果表明:改进后的小波去噪方法能够有效的去除信号中地磁噪声,从而提高测量数据的精度。目标定位算法研究。铁磁性目标能够导致其空间地磁场分布发生变化,形成磁异常。.人们通过分析和处理相应的磁异常,可以实现对磁性目标的定位。目前现有的定位算法中需事先已知目标体的类型,但在对未知目标实施定位时,这一条件很难满足。因此,本文提出了一种基于小子域的定位方法。该定位方法来源于小子域滤波算法,通过改进使其能够在不依赖于任何先知条件下,通过对磁异常自动压缩的方式确定磁性物体的位置。通过理论仿真,结果表明基于小子域的定位方法能够准确的确定磁性物体的位置。磁性目标定位的实验。为验证以上理论研究的正确性与有效性,本文开展了磁性目标定位的实验。实验表明:改进后的小波去噪方法可以提高地磁信号的信噪比,从而提高了地磁数据的精度;利用去噪后的地磁数据构建该区域的泰勒多项式地磁场模型,计算出高精度的目标磁异常;通过改进后的克里金插值方法,获得磁异常的空间分布;通过小子域定位算法对空间磁异常的数据进行处理,实现了对磁性目标的定位。
[Abstract]:In recent years, with the improvement of detection precision and detection speed of geomagnetic sensors, a large number of high-precision geomagnetic data can be obtained, which makes geomagnetic detection have a broader application prospects in various fields. Because of the existence of ferromagnetic target, its inherent magnetic field and induced magnetic field will be superimposed on the geomagnetic field, which will lead to the change of geomagnetic field distribution in space, and then produce magnetic anomaly in the space. Therefore, it is of great academic and practical significance to study how to use magnetic anomalies to locate magnetic targets. High-precision magnetic measuring instruments (optical pump magnetometer, superconducting quantum interferometer, etc.) have been widely used in geomagnetic detection at home and abroad. These instruments have high measuring accuracy and stability, and can be continuously measured for However, there are still some important technical difficulties in the process of using magnetic anomalies to locate targets, such as the extraction of magnetic anomalies, the spatial interpolation of geomagnetic maps, the method of noise suppression and location algorithm in the process of measurement. The research work mainly focuses on the following aspects: the extraction of magnetic anomaly is the key task of magnetic target localization. The geomagnetic field can be divided into two parts: normal field and anomalous field. The anomalous field is relative to the normal field, which is caused by the existence of magnetic target. In order to obtain accurate magnetic anomalies, it is necessary to construct a reasonable geomagnetic normal field model. The geomagnetic field model is divided into global model and regional model. The accuracy of target magnetic anomaly extraction is analyzed in this paper. Two geomagnetic anomaly models, polynomial model and Spline model, are constructed. The advantages and disadvantages of the two models are compared. The problems of the two models are analyzed, and the measures to improve the accuracy and the best model selection scheme are given. Because it is difficult to achieve large-scale geomagnetic survey conditions and obtain sufficient measurement data in actual surveys, it is necessary to use spatial interpolation method to improve the density of measurement data, thereby improving the accuracy of magnetic anomalies. In this paper, the Kriging interpolation method is used as an interpolation method in geomagnetic mapping. Based on the mathematical principle, the mathematical principle of the interpolation method is discussed. The variation function in the Kriging interpolation square can reflect the change of the local range and the characteristic in the specific direction. The variance function in the interpolation process is the fitting result of the experimental and theoretical variance functions. In this paper, a variance function fitting method based on particle swarm optimization is proposed, which can effectively improve the fitting accuracy of the variance function and thus improve the accuracy of geomagnetic map drawing. In order to improve the quality of the data, it is necessary to eliminate the noise in the geomagnetic signal. Because signals and noises have different forms in wavelet domain, wavelet de-noising can separate noises by their amplitudes varying with scales. In this paper, some shortcomings of traditional threshold function in wavelet threshold de-noising are pointed out. A threshold function which can dynamically adjust the threshold value is proposed to avoid the invariable attenuation of the wavelet coefficients with large absolute value in soft threshold method. The real results show that the improved wavelet de-noising method can effectively remove the geomagnetic noise in the signal and improve the accuracy of the measurement data. Research on target location algorithm. The existing localization algorithms need to know the type of target in advance, but it is difficult to satisfy this condition when locating unknown target. Therefore, this paper proposes a localization method based on small subdomain. This localization method is derived from small subdomain filtering algorithm and can be improved to be independent of any prophet. In order to verify the correctness and validity of the above theoretical research, the magnetic object location is carried out in this paper. Experiments show that the improved wavelet de-noising method can improve the signal-to-noise ratio of geomagnetic signals and thus improve the accuracy of geomagnetic data; the Taylor polynomial geomagnetic field model of this region is constructed by using the de-noised geomagnetic data, and the high-precision target magnetic anomaly is calculated; the space of magnetic anomaly is obtained by the improved Kriging interpolation method. Distribution between magnetic objects and magnetic anomaly data are processed by small subdomain localization algorithm.
【学位授予单位】：哈尔滨工程大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：P318.63;TP212.9
[Abstract]:In recent years, with the improvement of detection precision and detection speed of geomagnetic sensors, a large number of high-precision geomagnetic data can be obtained, which makes geomagnetic detection have a broader application prospects in various fields. Because of the existence of ferromagnetic target, its inherent magnetic field and induced magnetic field will be superimposed on the geomagnetic field, which will lead to the change of geomagnetic field distribution in space, and then produce magnetic anomaly in the space. Therefore, it is of great academic and practical significance to study how to use magnetic anomalies to locate magnetic targets. High-precision magnetic measuring instruments (optical pump magnetometer, superconducting quantum interferometer, etc.) have been widely used in geomagnetic detection at home and abroad. These instruments have high measuring accuracy and stability, and can be continuously measured for However, there are still some important technical difficulties in the process of using magnetic anomalies to locate targets, such as the extraction of magnetic anomalies, the spatial interpolation of geomagnetic maps, the method of noise suppression and location algorithm in the process of measurement. The research work mainly focuses on the following aspects: the extraction of magnetic anomaly is the key task of magnetic target localization. The geomagnetic field can be divided into two parts: normal field and anomalous field. The anomalous field is relative to the normal field, which is caused by the existence of magnetic target. In order to obtain accurate magnetic anomalies, it is necessary to construct a reasonable geomagnetic normal field model. The geomagnetic field model is divided into global model and regional model. The accuracy of target magnetic anomaly extraction is analyzed in this paper. Two geomagnetic anomaly models, polynomial model and Spline model, are constructed. The advantages and disadvantages of the two models are compared. The problems of the two models are analyzed, and the measures to improve the accuracy and the best model selection scheme are given. Because it is difficult to achieve large-scale geomagnetic survey conditions and obtain sufficient measurement data in actual surveys, it is necessary to use spatial interpolation method to improve the density of measurement data, thereby improving the accuracy of magnetic anomalies. In this paper, the Kriging interpolation method is used as an interpolation method in geomagnetic mapping. Based on the mathematical principle, the mathematical principle of the interpolation method is discussed. The variation function in the Kriging interpolation square can reflect the change of the local range and the characteristic in the specific direction. The variance function in the interpolation process is the fitting result of the experimental and theoretical variance functions. In this paper, a variance function fitting method based on particle swarm optimization is proposed, which can effectively improve the fitting accuracy of the variance function and thus improve the accuracy of geomagnetic map drawing. In order to improve the quality of the data, it is necessary to eliminate the noise in the geomagnetic signal. Because signals and noises have different forms in wavelet domain, wavelet de-noising can separate noises by their amplitudes varying with scales. In this paper, some shortcomings of traditional threshold function in wavelet threshold de-noising are pointed out. A threshold function which can dynamically adjust the threshold value is proposed to avoid the invariable attenuation of the wavelet coefficients with large absolute value in soft threshold method. The real results show that the improved wavelet de-noising method can effectively remove the geomagnetic noise in the signal and improve the accuracy of the measurement data. Research on target location algorithm. The existing localization algorithms need to know the type of target in advance, but it is difficult to satisfy this condition when locating unknown target. Therefore, this paper proposes a localization method based on small subdomain. This localization method is derived from small subdomain filtering algorithm and can be improved to be independent of any prophet. In order to verify the correctness and validity of the above theoretical research, the magnetic object location is carried out in this paper. Experiments show that the improved wavelet de-noising method can improve the signal-to-noise ratio of geomagnetic signals and thus improve the accuracy of geomagnetic data; the Taylor polynomial geomagnetic field model of this region is constructed by using the de-noised geomagnetic data, and the high-precision target magnetic anomaly is calculated; the space of magnetic anomaly is obtained by the improved Kriging interpolation method. Distribution between magnetic objects and magnetic anomaly data are processed by small subdomain localization algorithm.
【学位授予单位】：哈尔滨工程大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：P318.63;TP212.9
【相似文献】
相关期刊论文 前10条
1 Tadashi Nkatsuka;赵维勇;;磁异常的面间变换[J];国外地质勘探技术;1983年07期
2 熊光楚;_5志，
本文编号：2192098
本文链接：https://www.wllwen.com/shoufeilunwen/xxkjbs/2192098.html
