抗差估计在自由落体式绝对重力仪中的应用

针对自由落体式绝对重力仪中的异常数据,提出抗差参数估计的数据处理方法。对比分析传统最小二乘拟合与抗差最小二乘拟合计算的重力值,结果表明,抗差参数估计通过合理调整权重因子抑制了异常数据的影响,提高了重力测量精度。     

一种选权迭代法的半参数回归模型在滑坡预测中的应用

传统抗差估计中的选权迭代法无法探测到模型的系统误差,而基于补偿最小二乘原理的半参数模型可以较好地分离出系统误差。通过建立选权迭代法的半参数回归模型,利用时间序列法、L曲线法分别确定模型中的正则化矩阵及平滑因子,并利用选权迭代法重新定权,同时降低了观测粗差和系统误差对参数估值的影响。通过仿真算例,并以重庆奉节县大坪滑坡实测数据为例,验证了选权迭代法的半参数回归方法应用到三峡库区滑坡预测的有效性和优越性。     

基于相空间重构的FOA-GLSSVM深基坑变形预测模型研究

针对LSSVM模型参数选择的随机性与单一变量序列高维度重构参数选择的困难性,将相空间重构理论、果蝇优化算法引入LSSVM模型中,建立基于相空间重构的FOA-GLSSVM变形预测模型。为了验证提出模型的有效性与可靠性,结合具体工程实例与GLSSVM、支持向量机模型及最小二乘支持向量机模型进行对比研究。结果表明,提出的模型精度更好、稳定性更强。     

基于多尺度相关性分析的InSAR对流层延迟误差改正算法

针对合成孔径雷达干涉技术中对流层延迟误差会影响DEM精度的问题,提出采用小波多尺度相关性分析方法来减弱与高程相关的对流层延迟误差的影响,来提高合成孔径雷达干涉DEM的估计精度。该方法基于小波多分辨率分析理论,根据差分干涉相位不同组成的频率特性,利用小波分解重构均方根误差变化率确定分解层数,降低地形残差相位、噪声相位等对大气延迟误差相位估计的干扰,提取对流层延迟误差相位所在频带;然后结合对流层延迟误差相位和雷达坐标系下的DEM在不同尺度上的相关性定权并进行降权处理,重构解缠差分干涉图,改正差分干涉相位中与高程相关的对流层延迟的影响。采用本文方法对覆盖河南义马地区的2景ENVISAT ASAR数据进行处理,得到对流层延迟误差改正后的差分干涉图,估计的与高程相关的对流层延迟相位,与地形变化情况吻合。将对流层延迟误差改正后的干涉图用于DEM高程估计,结果显示本文方法重建的DEM与Aster GDEM的标准差由30.7 m提高到26.37 m,提高了InSAR DEM估计精度。     

LaCoste & Romberg ET型重力仪数字静电反馈系统的研制

介绍以计算机技术和采用数字比例积分微分控制器控制算法研制的数字式静电反馈系统的工作原理,并将该系统用于LaCoste & Romberg(LCR)ET20型重力仪的升级改造中.改造后LCR ET20重力仪的观测结果表明,该系统能解决模拟静电反馈系统中线性度不易调整和长期稳定度差等问题,并能提高重力仪的观测精度.     

土地调查中有关城镇与农村土地面积问题的研究

第二次全国土地调查工作中,城镇与农村地区的土地面积在计算方法上有所不同：农村地区的土地面积是以椭球面为基准的面积,即球面面积;城镇地区的土地面积是以平面为基准的面积,即平面面积。二者之间存在差异并按一定规律变化,当调查区处于投影带边缘附近时,同一图斑的球面面积与平面面积之间的差异达到最大。由于采用不同的计算方法而导致面积数量的差异,在进行农村与城镇有关土地数据整合时会出现矛盾。为了建立城乡统一的土地管理信息系统,保证土地面积的一致性,应当采取相应措施消除二者之间的差异。     

Relationships between hydrodynamic parameters and grain size in two contrasting transitional environments: The Lagoons of Venice and Cabras, Italy

A comparison was made of shallow water sediments from the Lagoon of Venice (LV) and the Lagoon of Cabras (LC), comparing depositional environments and exploring the relationships between hydrodynamics and sedimentological parameters. The two water bodies are very different in size (LV: 360 km²; LC: 22 km²), and the sediments predominantly consist of silty-clay (LV: Mz ≈ 26 μm; LC: Mz ≈ 6 μm). However, there are large differences between the two lagoons with respect to sand (LV: mean 19%; LC: mean ~ 3%) and clay (LV: mean 20%; LC: mean 45%) contents. The Lagoon of Venice (mean depth ~ 1 m) can be considered a tidal basin, whereas the Lagoon of Cabras (mean depth ~ 2 m) has the character of a coastal lake in which wind is the main hydrodynamic forcing factor. A comparison of sediment grain-size distributions with water circulation patterns in different parts of the lagoons highlighted some interesting differences. Grain-size analyses of samples reveal a deficiency of particles around 8 μm in the LC, which is interpreted as reflecting the transition between cohesive flocs/aggregates and non-cohesive coarser silt particles, while the transition limit in the LV is ~ 20 μm. Thus, particles are cohesive below 8 μm in the LC and below ~ 20 μm in the LV. This is probably because of the differences in the clay/silt ratio, which is much lower in the LV (~ 0.3) than in LC (~ 1), conferring a “silt-dominated network structure” on most of the LV sediments.The hydrographical data used were root mean square velocity (RMSV) and water residence time (WRT), computed under the main forcing conditions. The results show a general correlation between RMSV and sortable silt in the LC, and between RMSV and coarser sediments (63–105 μm) in the LV. Some significant differences between the lagoons were detected in the degree of correlation between WRT and grain size. Root mean square velocity (~ 7 cm s^− 1 in the LV and ~ 3 cm s^− 1 in the LC) was a greater forcing factor in the LC than in the LV. Conversely, WRT, which is on average ~ 16 days in the LV and ~ 19 days in the LC, has more influence in the LV. This study highlights the usefulness of comparing environments with different hydrodynamic energies, e.g., tidal and/or wind-driven currents, to elucidate and thereby improve our understanding of the processes governing the spatial distribution of sedimentological features, the transport mechanisms of sediments, and the relationship between them. The results demonstrate that the approach outlined in this study has the potential to provide a universal hydro-sedimentological classification scheme.     

用抽水试验资料确定水文地质参数的新方法

叙述了用最小二乘逼近法和最优化基本原理确定水文地质参数的新方法。在方程式的迭代过程中，水文地质参数的正确性是由水位降深值和试验测量值拟合结果的平均相对误差进行评价的。     

河西走廊黑河鼎新至哨马营段河水与地下水转化途径分析

运用美国ＧＥＯＭＥＴＲＩＣＳ公司制造的ＳｔｒａｔａｇｅｃｍＥＨ４电导率成像系统。对黑河鼎新至哨马营河谷地带进行实地调查发现,此段存在一地堑式断层,为东西走向,古道沿此断层形成。河流在此段大部分沿古河道为地下水,自西向东流去,在东部板滩井一带（盐碱沼泽地）以垂向蒸发方式通过地表及植被排泄。古尔乃绿洲的形成与黑河河水通过断层转化成地下水有关。     

利用时域有限差分方法计算雷电水平电场

雷电水平电场是计算电力架空线雷电感应过电压的重要参数,为了准确获得该参数值,本文利用时域有限差分方法(FDTD)对不同地面电导率情况以及距雷电回击通道不同距离处的雷电水平电场进行了计算和分析,结果表明:(1)地面电导率均匀情况下,地面电导率越大,计算距离越近,雷电水平电场波形愈加接近模式输入的电流波形,这与雷电水平电场中的辐射场分量较小有关,而增大地面电导率或者增加计算距离都会导致计算结果变小甚至波形出现双极性的特征;(2)地面电导率水平分层情况下,计算得到的雷电水平电场受上层土壤的电导率影响更大,但是下层不同电导率的土壤一定程度上改变了土壤整体的等效电导率,使得计算结果与均匀电导率情况下有所差别。