军用数字地图交通要素实体化算法

数字地图中描述地理要素的方法与现实世界中地理实体的存在方式并不吻合,这种描述与现实的不一致对地理信息系统(GIS)中进行一些较为复杂的查询和应用造成障碍。实体化的方法就是要缩小描述与现实的差距,使地理信息按照"实体"的方式组织。本文介绍了军用数字地图交通要素实体化的思想以及算法的实现步,探讨了利用军用基础地理信息获取战场环境专题数据的方法。     

Thermochemistry of monazite-(La) and dissakisite-(La): implications for monazite and allanite stability in metapelites

Thermochemical properties have been either measured or estimated for synthetic monazite, LaPO₄, and dissakisite, CaLaMgAl₂(SiO₄)₃OH, the Mg-equivalent of allanite. A dissakisite formation enthalpy of ?6,976.5 ± 10.0 kJ mol^?1 was derived from high-temperature drop-solution measurements in lead borate at 975 K. A third-law entropy value of 104.9 ± 1.6 J mol^?1 K^?1 was retrieved from low-temperature heat capacity (C _p) measured on synthetic LaPO₄ with an adiabatic calorimeter in the 30–300 K range. The C _p values of lanthanum phases were measured in the 143–723 K range by differential scanning calorimetry. In this study, La(OH)₃ appeared as suitable for drop solution in lead borate and represents an attractive alternative to La₂O₃. Pseudo-sections were calculated with the THERIAK-DOMINO software using the thermochemical data retrieved here for a simplified metapelitic composition (La = ∑REE + Y) and considering monazite and Fe-free epidotes along the dissakisite-clinozoïsite join, as the only REE-bearing minerals. Calculation shows a stability window for dissakisite-clinozoïsite epidotes (T between 250 and 550°C and P between 1 and 16 kbar), included in a wide monazite field. The P–T extension of this stability window depends on the bulk-rock Ca-content. Assuming that synthetic LaPO₄ and dissakisite-(La) are good analogues of natural monazite and allanite, these results are consistent with the REE-mineralogy sequence observed in metapelites, where (1) monazite is found to be stable below 250°C, (2) around 250–450°C, depending on the pressure, allanite forms at the expense of monazite and (3) towards amphibolite conditions, monazite reappears at the expense of allanite.     

The compressibility of Fe- and Al-bearing phase D to 30 GPa

High-pressure in situ X-ray diffraction experiment of Fe- and Al-bearing phase D (Mg_0.89Fe_0.14Al_0.25Si_1.56H_2.93O₆) has been carried out to 30.5 GPa at room temperature using multianvil apparatus. Fitting a third-order Birch–Murnaghan equation of state to the P–V data yields values of V ₀ = 86.10 ± 0.05 ų; K ₀ = 136.5 ± 3.3 GPa and K′ = 6.32 ± 0.30. If K′ is fixed at 4.0 K ₀ = 157.0 ± 0.7 GPa, which is 6% smaller than Fe–Al free phase D reported previously. Analysis of axial compressibilities reveals that the c-axis is almost twice as compressible (K _c  = 93.6 ± 1.1 GPa) as the a-axis (K _a  = 173.8 ± 2.2 GPa). Above 25 GPa the c/a ratio becomes pressure independent. No compressibility anomalies related to the structural transitions of H-atoms were observed in the pressure range to 30 GPa. The density reduction of hydrated subducting slab would be significant if the modal amount of phase D exceeds 10%.     

New high-pressure phases in BaCO<Subscript>3</Subscript>

Barium carbonate (BaCO₃) was examined in a diamond anvil cell up to a pressure of 73 GPa using an in situ angle-dispersive X-ray diffraction technique. Three new phases of BaCO₃ were observed at pressures >10 GPa. From 10 to 24 GPa, BaCO₃-IV had a post-aragonite structure with space group Pmmn. There are two molecules in a single unit cell (Z = 2) of the orthorhombic phase, which is same as the high-pressure phases of CaCO₃ and SrCO₃. The isothermal bulk modulus of BaCO₃-IV is K ₀ = 84(4) GPa, with V ₀ = 129.0(7) Å³ when K ₀′ = 4. The c axis of the unit cell parameter is less compressible than the a and b axes. The relative change in volume that accompanies the transformation between BaCO₃-III and BaCO₃-IV is ～6%. BaCO₃-V, which has an orthorhombic symmetry, was synthesized at 50 GPa. As the pressure increases, BaCO₃-V is transformed into tetragonal BaCO₃-VI. This transformation is likely to be second order, because the diffraction pattern of BaCO₃-V is similar to that of BaCO₃-VI, and some single peaks in BaCO₃-VI become doublets in BaCO₃-V. After decompression, the new high-pressure phases transform into BaCO₃-II. Our findings resolve a dispute regarding the stable high-pressure phases of BaCO₃.     

三维激光扫描技术与断层面形貌量化分析在基岩区古地震研究中的应用

由于对第四纪地层的严重依赖,传统古地震探槽研究方法在基岩区难以发挥作用,导致无法获取基岩区断层的强震活动历史。本研究以山西地堑系的交城断裂为目标断裂,以断裂北段2处基岩断层面为研究对象,通过三维激光扫描技术获取基岩断层面高精度形貌,基于变差函数法结合滑动窗口操作量化断层表面形貌特征,开展在基岩区提取断裂古地震信息的实例研究。结果显示,2处基岩断层面的形貌在高度上具有明显的分段特征,指示了断层面在地震事件作用下的分段出露过程。这种断层面形貌分段特征可以用来识别古地震事件和同震位移量。在思西村基岩断层面上,识别出由老到新的3次古地震事件,同震倾滑位移量依次为2.0 m、1.9 m和2.3 m,在上兰镇基岩断层面上,识别出由老到新的3次古地震事件,同震倾滑位移量依次为1.4 m、2.5 m和2.0 m,指示了交城断裂北、中段具有产生同震位移量大于2 m、震级大于7.5级的破裂型地震的能力。上述研究成果表明,基于三维激光扫描和形貌量化分析方法开展基岩断层面古地震研究,可以准确而高效地识别古地震事件次数和同震位移量,扩展古地震的研究对象,拓宽古地震的研究空间。在未来的研究中,可以适时地开展宇宙成因核素测年以测定断层面的暴露年龄,获得发震年代,给予地震序列年龄框架。     

环境激励下基于波动能量法的结构损伤识别

目前作为结构健康监测系统核心的损伤识别大多是基于模态参数变化而进行的,但模态参数对局部损坏不敏感,导致损伤识别精度不够。波在结构中的传播状态可以更好地反映局部损伤状况,波动能量可以作为损伤识别的有效指标。为了提高环境激励下结构损伤识别的精度,采用S变换分析了结构输出信号,建立波动能量指标,从而使波动能量指标的使用领域扩展到非平稳信号范围。最后通过三层钢框架试验及弹性分层剪切梁的数值模型对该方法进行了验证,结果表明:该方法不仅能够有效识别结构损伤位置,而且能够识别出损伤程度。     

3种经典机器学习算法在火山岩测井岩性识别中的对比

岩性识别一直是火山岩油气勘探中的重要课题,基于测井数据的岩性识别也逐渐成为火山岩研究的需要,大数据背景下的机器学习算法为测井岩性识别提供了一个新方向。为提高某研究区火山岩岩性识别符合率,本文采用K近邻、支持向量机和自适应增强3种经典机器学习算法,对研究区内的粗面岩、非致密粗面岩、辉绿岩、辉长岩、玄武岩和非致密玄武岩等6类中基性火山岩进行岩性识别。从常规测井系列中优选对研究区岩性敏感的自然伽马、声波时差、补偿中子、深侧向电阻率和补偿密度等5种测井参数作为岩性识别模型的输入向量,从研究区内5口有岩心样品或薄片鉴定资料的目标层中选取测井数据点1 440个,其中960个作为训练样本,其余480个作为测试样本。以识别符合率和时间作为评价指标,对3种算法的识别结果进行对比分析,实验表明：自适应增强算法的分类准确率最高,6类岩性平均识别符合率达到82.10%;支持向量机算法表现良好,平均识别符合率为81.04%;K近邻算法平均识别符合率为76.04%。     

XRF半定量分析技术在矿石光片鉴定中的应用

偏光显微镜反光镜下鉴定金属矿物是最常用、最快捷有效的方法,然而鉴定结果容易受到鉴定者主观因素或磨片情况的影响。为了更快速、准确地在反光镜下鉴定矿物,利用XRF半定量分析技术快速、无损、前处理简单等特点,以及磨平抛光的矿石矿物在半定量分析过程中能较好地避免楔入效应和阴影效应的有利因素,采取了先用XRF半定量分析测得的结果去反推矿物,再在显微镜下鉴定矿物的方法。以多金属硫化物矿石、含铀多金属矿石及含微量铀的矿石样品为例进行鉴定,结果表明,在鉴定前知道光片上的元素及其质量分数,通过地球化学方法和经验法反推矿物,结合显微镜观察需要鉴定的矿物,可有效避免忽略掉透明/半透明矿物（锡石）、小粒径矿物（沥青铀矿）、易混淆矿物（铀黑）,提高光片鉴定工作的效率和准确性。     

伪随机源时间域航空电磁响应系统辨识

在时间域航空电磁系统中,采用高频伪随机序列作为激励源是实现浅层高分辨地质探测的有效途径。为了提高大地脉冲响应的辨识精度,在基本相关辨识方法的基础上,本文提出了一种基于时间域循环相关计算的高精度辨识方法,通过消除旁瓣效应影响和在时间域下求解维纳-霍夫方程获取大地脉冲响应。数值模拟仿真结果表明,辨识大地响应相对误差均方根为2.2×10^-7,早期大地响应辨识误差小于10^-10。该高精度辨识方法具有较好的噪声压制能力,在观测时间较短时,单周期辨识提升的信噪比为8.7 dB,优于同等条件下线性叠加降噪方法提升的信噪比6.1 dB。     

基于岩样细观图像深度学习的岩性自动分类方法

在现阶段的岩土工程中,通常采用人工识别的方法来判别岩样种类,不仅耗时长、专业性强,还易受主观因素影响,准确率不理想。随着计算机技术的发展,机器学习逐渐被应用于岩性的自动识别,开启了岩样分类的新路径。本文以重庆市主城区4种典型岩样（泥岩、砂质泥岩、泥质砂岩和砂岩）的细观图像为研究对象,基于Inception V3卷积网络模型和迁移学习算法,建立了岩样细观图像深度学习模型,并完成了训练学习。结果显示：模型在训练1 000次后,训练集中的分类准确率达到92.77%,验证集中的分类准确率为76.31%。其中,验证集中的砂岩识别准确率为97.28%,泥岩识别准确率为81.85%,泥质砂岩识别准确率为72.59%,砂质泥岩识别准确率为72.35%。与现有的机器学习方法相比,本识别模型不仅可以自动识别岩性极为相近的岩样,而且具有较好的识别准确率、鲁棒性和泛化能力。