钢管混凝土柱带钢板耗能键组合剪力墙抗震研究

提出了连排钢管混凝土柱带钢板耗能键组合剪力墙,它由钢管混凝土连排柱、柱间钢板耗能键、钢板耗能键外包混凝土条带三种单元组合而成。进行了4个不同设计参数试件的低周反复荷载试验研究。分析了各试件的承载力、刚度及其退化过程、滞回特性、延性和破坏特征,探讨了分灾耗能机制。研究表明:连排钢管混凝土柱带钢板耗能键组合剪力墙,承载力较大,后期刚度较稳定;混凝土条带在开裂与闭合过程中消耗地震能量,钢板耗能键通过弯剪变形消耗地震能量,钢管与混凝土条带共同工作协同耗能,具有良好的抗震耗能机制;这种新型组合剪力墙具有较强综合抗震耗能能力。     

条形基础下水平布置纤维增强聚合物筋材地基试验研究

纤维增强聚合物(FRP)具有耐腐蚀、耐水、耐高温、高强度、高弹性模量、质量轻等优点,将其用作加筋材料不仅能显著提高加筋地基的加筋效果,还能有效解决传统加筋材料的耐久性问题。分别对纯砂和5种水平加筋方式进行了加筋地基模型试验,测试了各级荷载下的地基沉降、FRP的应变及土压力数值,并初步探讨了FRP加筋的作用机制。试验结果表明,加筋能显著提高地基承载力和减小地基沉降,特别是双层水平加筋方式的加筋效果更加明显,而锚固件的设置对加筋效果的影响并不大,水平FRP筋材主要起到了土压力扩散作用。     

潮汐效应对沿海多条带时序InSAR地表形变监测的影响

潮汐效应对沿海大范围、高精度的合成孔径雷达干涉测量(Interferometric Synthetic Aperture Radar, InSAR)技术地表形变监测的影响不可忽略. 潮汐效应中的固体潮(Solid Earth Tide, SET)位移和海洋潮汐负荷(Ocean Tide Loading, OTL)位移会在时间和空间域上对沿海大范围InSAR地表形变监测产生分米级的误差, 而且不同条带中地表潮汐形变时空差异性会对多条带InSAR影像拼接产生较大误差. 本文针对沿海地区多条带InSAR形变时序, 重点分析了潮汐效应在不同条带InSAR地表沉降监测中的影响, 并采用多种潮汐位移估计方法对多条带地表潮汐形变进行改正. 结果表明, 研究区域不同条带时序InSAR中地表潮汐形变具有较大的时空差异性, 其造成沿海地区相邻条带时序InSAR地表沉降速率差异能达到1~2 cm·a^-1; 通过固体潮模型、海潮负荷模型或GPS参考站网海潮位移改正, 能够消除不同条带中地表潮汐形变时空差异性对时序InSAR形变结果拼接产生的空间高阶非线性误差, 弥补了传统拼接方法在拟合潮汐位移偏差的不足. 在大气延迟误差改正的基础上, 时序InSAR形变残差的标准差(Standard Deviation, STD)由潮汐改正前的2.3 cm减少至0.75 cm; 本文研究揭示了沿海地表潮汐形变改正对多条带时序InSAR形变结果拼接的重要性, 可大幅提高沿海地区广域时序InSAR地表形变监测的精度.

     

Vertical vibration of a rigid strip footing on viscoelastic half-space

This paper presents an analytical method for modeling the dynamic response of a rigid strip footing subjected to vertical-only loads. The footing is assumed to rest on the surface of a viscoelastic half-space; therefore, effects of hysteretic soil damping on the impedance of the foundation and the generated ground vibrations are considered in the solution. To solve the mixed boundary value problem, we use the Fourier transform to cast a pair of dual integral equations providing contact stresses, which are solved by means of Jacobi orthogonal polynomials. The resulting soil and footing displacements and stresses are obtained by means of the Fourier inverse transform. The solution provides more realistic estimates of footing impedance, compared to existing solutions for elastic soil, as well as of the attenuation of ground vibrations with distance of the footing. The latter is important for the estimation of machine vibration effects on nearby structures and installations.     

Optimisation of underwater visual census and controlled angling methods for monitoring subtidal temperate reef fish communities

Standardised sampling protocols for monitoring fish stocks are essential to assess changes in stock status and provide a means to evaluate the effectiveness of fisheries management measures, such as marine protected areas (MPAs). This study aimed to optimise two standard sampling methods, namely underwater visual census (UVC) and controlled angling, for assessing subtidal reef fish communities. In terms of efficiency, variability and bias, UVC transects were found to be superior to point counts. For controlled angling, an effort of two angler-hours per fishing station provided low catch variability, high catch per unit effort and a representative catch. Whereas UVC provided less variable estimates of relative density, controlled angling provided greater sampling efficiency. It is thus recommended that the two methods be used in conjunction. The optimal sampling protocols identified are suitable for rapid assessments or long-term monitoring of subtidal, temperate reef fish communities.     

基于Transformer增强型U-net的CT图像稀疏重建与伪影抑制

实现低剂量计算机断层成像（CT）的一个有效办法是减少投影角度,但投影角度较少会产生严重的条状伪影,降低图像的临床使用价值。针对该问题,提出一种耦合卷积神经网络（CNN）和多种注意力机制的U型网络（TE-unet）。首先采用U型架构提取多尺度特征信息;其次提出一个包含CNN和多种注意力的模块提取图像特征;最后在跳跃连接处加入Transformer块过滤信息,抑制不相关特征,突出重要特征。所提网络结合CNN的局部特征提取能力和Transformer的全局信息捕获能力,辅以多种注意力机制,实现了良好的去条状伪影能力。在60个投影角度下,与经典的Uformer网络相比,峰值信噪比（PSNR）高出0.3178 dB,结构相似度（SSIM）高出0.002,均方根误差（RMSE）降低0.0005。实验结果表明,所提TE-unet重建的图像精度更高,图像细节保留的更好,可以更好地压制条状伪影。

     

条带开采地表沉陷预计参数的确定

概率积分法是条带开采地表移动和变形预计常用的方法,而概率积分法预计的精度取决于其预计参数的确定。以国内大量的条带开采实测资料为基础,应用相似理论对条带开采地表沉陷的相似现象进行了模糊聚类分析,计算出了条带开采地表沉陷预计参数。根据所得条带开采地表沉陷相似现象的分类及其地表沉陷预计参数,应用模式识别对待求条带开采地表沉陷预计参数进行了求取。工程实例表明,用模糊优化确定的预计参数进行条带开采地表移动和变形预计,其预计结果更加可靠、准确。     

A test method for analyzing the deformation of landslide model

In order to study the overall deformation of geotechnical model conveniently, the worksite of landslide bridge foundation reinforced by the front and rear row anti-slide piles in Chenglan railway was taken as an example. On the basis of shaking tabe test of a 1/40 reduced scale model, the landslide deformation caused by vibration waves was monitored through burying self-made phosphor bronze strips in soil. Combined with the horizontal and vertical coordinates of the bending strain points on the phosphor bronze strips, the digital matrix was converted by applying Renka Cline random matrix generation method, and the two-dimensional contour plots were drawn based on it. The results showed that the two-dimensional contour plots reflected the basic law of landslide deformation reasonably, and it revealed the evolution process of landslide deformation and failure. The research conclusions were consistent with the test phenomenon, which met the basic requirements of overall deformation analysis of landslide model. This proposed method can monitor multiple cross sections and was practical for model test.     

Seismic bearing capacity of shallow strip footings

Seismic bearing capacity of shallow strip footings in soil has been obtained in the form of pseudo-static seismic bearing capacity factors N_cd, N_qd and N_d, denoting the cohesion, surcharge and unit weight components, respectively, by an extensive numerical iteration technique. Limit equilibrium method of analysis with composite failure surface is assumed. The validity of the principle of superposition is examined. Effects of both the horizontal and vertical seismic acceleration coefficients have been found to always reduce the ultimate bearing capacity significantly. Results obtained by the present method of analysis are compared with the available results and are found to be the least in the seismic case.     

一种机载LiDAR条带平差最小二乘匹配算法

基于七参数正形变换的数据驱动模型实现了机载LiDAR条带平差,算法借鉴了Robert(2004)的最小二乘表面匹 配思想,通过引入高斯–马尔科夫模型改进了原有算法,得到未知参数的最小无偏方差估计。实验采用两组实测数 据,分别考察了引入高斯–马尔科夫模型的必要性、算法效率以及迭代收敛性和算法精度。实验表明：(1)剖面检查吻 合且精度一致;(2)TerraMatch量测匹配精度,理想数据高程匹配误差小于0.05 m,数据质量不理想时误差稍大,但均 能成功匹配。