τ-P域地震资料三维叠前深度偏移及其实用算法

本文提出了在τ-P域实现三维叠前深度偏移的计算方法，该方法把地震道集由时-空域变换到τ-P域，然后对共P数据体进行相移或相移加插值偏移，将偏移后的各共P数据体叠加得到叠前深度偏移结果．该算法能够实现陡倾地层的正确归位，降低偏移处理维数，减少计算机内存需求，易于并行处理，但是计算量和所花费的计算机时间仍然很大．为此我们又提出了针对目的层的分时（层）偏移方法，由于只考虑目的层，可节省大量的计算时间．文中以二维实际地震资料及三维数值模型资料的处理说明了上述方法的可行性与有效性．     

综合分析技术在微细粒浸染型金矿金赋存状态研究中的应用

采用化学组分分析、物相分析、粒度筛析、X射线面扫描分析、矿物分离分析并结合数理统计分析等综合分析技术对微细粒浸染型金矿进行工艺矿物学研究,较为快速、准确地查明该金精矿金的赋存状态.几种分析方法互为补充,结果吻合.     

水质大涡模拟数学模型研究

建立了基于弱可压缩流基础上的水质大涡模拟数学模型及污染物质扩散系数与平均运动的关系。数值计算采用有限体积法和预测校正法,固壁边界条件采用"部分滑移条件"。并应用该模型对重庆嘉陵江、长江两江交汇段的流场,污染物质浓度分布进行了计算,计算结果与实测值比较吻合。研究表明,横向扩散与流动状态有关,当支流流量相对较小时,横向扩散较弱,形成岸边污染带,当支流流量相对较大时,横向扩散较强,污染带向江心移动。     

降雨条件下岩土饱和-非饱和渗流分析

基于Buckley-Leverett两相渗流方程,提出了新的岩土饱和-非饱和渗流数学模型,利用有限差分方法给出了隐式压力显示饱和度数值求解方法,编制了饱和-非饱和渗流计算程序.结合工程实例,模拟了降雨入渗条件下边坡岩体渗流场孔隙压力变化和含水饱和度变化,模拟结果验证了所提出的模型对饱和-非饱和渗流的有效性.     

面向数字化测绘生产工作流的Web存储系统

讨论了一个面向工作流的项目管理系统中 ,各类计划、实施、监督管理、汇总报表等结构化文档及非结构化文档的存储访问系统的构造方法 ,并设计与实现了一个数字化测绘生产管理系统实例     

人口生态生产:人地作用机制下"三江并流"世界遗产保护的新视角

基于不同人口需要而进行的人类活动对“三江并流”区的地理环境产生了巨大影响,形成了独具特色的人地作用机制。这一作用机制一方面对“三江并流”世界遗产的保护构成了威胁,但另一方面,通过人口生态生产,实现人口数量、质量、结构和分布的生态化,改进人类活动,也可优化“三江并流”区的地理环境,保护“三江并流”世界遗产,实现该地区的可持续发展。     

Variability in the vertical hyporheic water exchange affected by hydraulic conductivity and river morphology at a natural confluent meander bend

River confluences and their associated tributaries are key morphodynamic nodes that play important roles in controlling hydraulic geometry and hyporheic water exchange in fluvial networks. However, the existing knowledge regarding hyporheic water exchange associated with river confluence morphology is relatively scarce. On January 14 and 15, 2016, the general hydraulic and morphological characteristics of the confluent meander bend (CMB) between the Juehe River and the Haohe River in the southern region of Xi'an City, Shaanxi Province, China, were investigated. The patterns and magnitudes of vertical hyporheic water exchange (VHWE) were estimated based on a one‐dimensional heat steady‐state model, whereas the sediment vertical hydraulic conductivity (K_v) was calculated via in situ permeameter tests. The results demonstrated that 6 hydrodynamic zones and their extensions were observed at the CMB during the test period. These zones were likely controlled by the obtuse junction angle and low momentum flux ratio, influencing the sediment grain size distribution of the CMB. The VHWE patterns at the test site during the test period mostly showed upwelling flow dominated by regional groundwater discharging into the river. The occurrence of longitudinal downwelling and upwelling patterns along the meander bend at the CMB was likely subjected to the comprehensive influences of the local sinuosity of the meander bend and regional groundwater discharge and finally formed regional and local flow paths. Additionally, in dominated upwelling areas, the change in VHWE magnitudes was nearly consistent with that in K_v values, and higher values of both variables generally occurred in erosional zones near the thalweg paths of the CMB, which were mostly made up of sand and gravel. This was potentially caused by the erosional and depositional processes subjected to confluence morphology. Furthermore, lower K_v values observed in downwelling areas at the CMB were attributed to sediment clogging caused by local downwelling flow. The confluence morphology and sediment K_v are thus likely the driving factors that cause local variations in the VHWE of fluvial systems.     

一种采用基础矩阵描述的核线影像生成方法

设计了一种采用基础矩阵的立体像对核线影像制作方法,由基础矩阵来描述影像间的相对方位关系,避免了相对定向过程中复杂的迭代过程。通过稀疏匹配获取高精度的同名点,由基础矩阵确定每组同名点对应的同名核线,根据核线必相交于核点的原理,采用最小二乘法确定左右影像的核点坐标;然后根据核点坐标完成影像间核线的快速映射,沿核线方向采用双线性内插法进行核线重采样,完成核线影像制作。采用多组影像进行核线影像制作实验,对生成的核线影像进行同名特征点提取与匹配,结果表明采用该方法制作的核线影像上下视差较小,验证了方法的有效性。     

历史数据和强化学习相结合的低频轨迹数据匹配算法

针对低频(采样间隔大于1min)轨迹数据匹配算法精度不高的问题,提出了一种基于强化学习和历史轨迹的匹配算法HMDP-Q,首先通过增量匹配算法提取历史路径作为历史参考经验库;根据历史参考经验库、最短路径和可达性筛选候选路径集;再将地图匹配过程建模成马尔科夫决策过程,利用轨迹点偏离道路距离和历史轨迹构建回报函数;然后借助强化学习算法求解马尔科夫决策过程的最大回报值,即轨迹与道路的最优匹配结果;最后应用某市浮动车轨迹数据进行试验。结果表明:本文算法能有效提高轨迹数据与道路匹配精度;本算法在1min低频采样间隔下轨迹匹配准确率达到了89.2%;采样频率为16min时,该算法匹配精度也能达到61.4%;与IVVM算法相比,HMDP-Q算法匹配精度和求解效率均优于IVVM算法,16min采样频率时本文算法轨迹匹配精度提高了26%。