TI介质局部角度域高斯束叠前深度偏移成像

各向异性射线理论基础上的局部角度域叠前深度偏移方法能够为深度域构造成像与基于角道集的层析反演提供有力支撑,但是对于复杂地质构造而言,高斯度叠前深度偏移在不失高效、灵活等特点的情况下,具有明显的精度优势.为此,本文研究局部角度域理论框架下的高斯束叠前深度偏移方法.为提高算法效率与实用性,文中讨论了一种从经典弹性参数表征的各向异性介质运动学和动力学射线方程演变而来的由相速度表征的简便形式,并提出了一种比较经济的各向异性高斯束近似合成方案.结合地震波局部角度域成像原理,讨论一种适合高斯束偏移的角度参数计算方法.国际上通用的理论模型合成数据试验表明：相比局部角度域Kirchhoff叠前深度偏移成像方法,本文方法具有更高的成像精度与抗噪能力,既适用于复杂构造成像,也可为TI介质深度域偏移速度分析与模型建立提供高效的偏移引擎.     

Practical modified scheme of linear shallow-water equations for distant propagation of tsunamis

A simple but practical numerical model describing a distant propagation of tsunamis is newly proposed by introducing an additional term to the existing modified scheme. The numerical dispersion of the proposed model is manipulated to replace the physical dispersion of the linear Boussinesq equations without any limitation. The new model developed in this study is applied to propagation of a Gaussian hump over a constant water depth and the predicted free surface displacements are compared with available analytical solutions. A very reasonable agreement is observed.     

Experimental study of stick-slip behaviour

Simple axi-symmetric uni-axial compression tests have been realized on dry loose samples of glass beads (diameters d: d=0.2 ± 0.05 mm, 0.75 ± 0.1 mm, or 3 mm) and on Hostun sand under small lateral confinement, σ₃<60 kPa, using different sample sizes. The experiments with the two smallest spheres (d=0.2 and 0.75 mm) exhibit stick-slips, which are characterized by (i) a rapid release Δq of the deviatoric stress q and by (ii) the strain Δε₁ separating two events. The samples which exhibit stick-slip also present a weakening of strength q(ε₁) as the rate of deformation dε₁/dt is increased. No stick-slip is generated during the first part of the q−ε₁ curve, i.e. when q grows fast with ε₁. Four different parameters helped us determine the statistics of Δq and Δε: the lateral pressure σ₃′, the rate of deformation dε₁/dt, the sample height H, and the diameter D. The statistics do not depend on rate history. They look like exponentials in small samples and/or in (large sample+fast dε₁/dt), and they look like Poissonian or Gaussian in (Large sample+small dε₁/dt). This change in statistics is attributed to a varying of triggering process starting from a single random event in small samples to multiple random events. We have interpreted this change of statistics as due to some finite size effect so that the representative elementary volume shall contain at least (200)³ grains. Localization of deformation is visible at the end of compression but cannot be detected from stick-slip statistics nor from q vs ε curve. Copyright © 2004 John Wiley & Sons, Ltd.     

基于EM和BIC的直方图拟合方法应用于遥感变化检测阈值确定

利用遥感图像进行变化检测时,确定"差异图像"上各变化类型的阈值非常关键.本文引入图像直方图拟合方法来确定变化阈值.首先通过基于变化向量分析方法,得到变化强度图像,然后假设该变化强度图像中的像元值符合混合高斯分布模型,利用期望最大(EM)算法和贝叶斯信息准则(BIC)求出最佳的混合高斯分布模型,拟合此时的图像直方图,最后利用贝叶斯判别准则确定出各变化类型的变化阈值.试验证明,这种方法是一种较为有效的自动确定变化阈值的方法.     

非高斯有色噪声影响下的高斯混合滤波算法

针对GNSS/SINS紧组合定姿定位系统在非高斯有色噪声环境下直接使用高斯混合扩展卡尔曼滤波（Gaussian mixture extend Kalman filter, GMEKF）出现的随机模型失配现象,提出一种顾及非高斯有色噪声影响的GMEKF优化算法。该算法在高斯混合模型对非高斯噪声近似的基础上,通过状态扩增和量测组差对非高斯噪声序列中的有色噪声成分进行白化处理,从精化随机模型的角度改善了GMEKF算法的性能。实验结果表明,在非高斯有色噪声环境下,相比于GMEKF,该算法对随机模型建模更准确,将其用于受非高斯有色噪声影响的定姿定位场合中可进一步提高导航解算精度。     

融合像素级和对象级的遥感图像变化检测方法

针对现有遥感影像变化检测方法常存在的检测结果破碎、虚检较多、对数据匹配要求高等问题。提出了一种融合像素级和对象级的遥感图像变化检测方法。利用光谱和纹理信息构建单高斯模型,在多尺度上进行像素级变化检测。然后,以像素级检测结果为种子区域,同时在变化前后影像上区域生长,融合生长结果提取变化对象。最后,依据检测需求对变化对象进行特征分类并滤除虚警。实验结果表明,该方法降低了虚检,保持了变化区域的结构完整性,在变化前后图像分辨率存在一定差别时仍有较高的检测精度。     

全波形激光雷达的波形优化分解算法

随着数据存储能力和处理速度的提高,三维激光扫描系统逐渐具备全波形采集和分析技术。为了从全波形数据中获得脉冲时间、幅度、脉宽以及多回波分布等综合信息,波形分解成为了全波形激光雷达数据处理的关键技术之一。针对LM算法在一定程度上依赖初值,而传统激光雷达数据处理容易遗漏部分重叠的返回波,本文提出了一种改进回波分量初值设定的算法来获取回波脉冲的位置、宽度和强度。针对一套自主研发的全波形记录激光雷达演示系统进行了波形分解试验,定性和定量分析结果验证了该方法的有效性、可靠性和准确性。     

A robust cylindrical fitting to point cloud data

Environmental, engineering and industrial modelling of natural features (e.g. trees) and man-made features (e.g. pipelines) requires some form of fitting of geometrical objects such as cylinders, which is commonly undertaken using a least-squares method that—in order to get optimal estimation—assumes normal Gaussian distribution. In the presence of outliers, however, this assumption is violated leading to a Gaussian mixture distribution. This study proposes a robust parameter estimation method, which is an improved and extended form of vector algebraic modelling. The proposed method employs expectation maximisation and maximum likelihood estimation (MLE) to find cylindrical parameters in case of Gaussian mixture distribution. MLE computes the model parameters assuming that the distribution of model errors is a Gaussian mixture corresponding to inlier and outlier points. The parameters of the Gaussian mixture distribution and the membership functions of the inliers and outliers are computed using an expectation maximisation algorithm from the histogram of the model error distribution, and the initial guess values for the model parameters are obtained using total least squares. The method, illustrated by a practical example from a terrestrial laser scanning point cloud, is novel in that it is algebraic (i.e. provides a non-iterative solution to the global maximisation problem of the likelihood function), is practically useful for any type of error distribution model and is capable of separating points of interest and outliers.     

高斯两步移动搜索法的模型研究——以上海市绿地可达性为例

合理的配置公共服务设施对城市的现代化建设至关重要。空间可达性是度量公共服务设施配置是否合理的方法之一。在诸多的研究方法中,高斯两步移动搜索法由于直观且运算简便,因而得到广泛应用。但该方法也有不足之处,本文运用格网GIS方法,以上海市的绿地空间可达性为例,对两步移动搜索法进行模型方法的改进研究,并对高斯两步移动搜索法和格网化的高斯两步移动搜索法进行对比分析,结果表明：后者可降低空间可达性的误差,提高可达性精度,对上海市绿地空间可达性反映更加真实、客观。如果选择合理的数据将其格网化,该方法也可以用于其他公共服务设施的空间可达性研究。     

川西海相气田高含硫天然气集输装置泄漏快速模拟方法

川西气区已经开发多年,海相气藏的开发已经成为气田发展的必然趋势。川西海相气井中高含H2S,高含硫集输装置发生泄漏事故将引发毒性气体扩散,毒性气体会对周围一定距离的公众造成损害。本文通过分析川西人口密集区高含硫天然气泄漏扩散分布规律,对高斯扩散模型参数选取方法进行优化,设计了一种快速分析评价方法,形成一套针对集输管道含硫天然气、净化厂脱硫装置含H2S天然气泄漏的扩散模拟程序,实现对气体泄漏扩散后果的快速模拟。以彭州1井为例,模拟分析不同地形参数条件下集输管线、净化厂脱硫装置发生泄露后在风速0.5 m/s、1.5 m/s时的扩散模拟情况,及泄漏扩散后硫化氢浓度1 000 ppm、100 ppm所能到的最远安全距离,为确定公众防护距离评价提供决策支持。