PLS法与隧道围岩稳定性分类

围岩稳定性是隧道工程支护设计的基本参数,采用偏最小二乘回归法,综合考虑多因变量影响因素,对隧道围岩稳定性分类进行回归建模,并取得较好效果.     

The development and analysis of quasi-linear map projections

Map projections are an essential component of coordinate systems used in applications such as surveying, topographic mapping, and engineering, and care needs to be taken to select ones that minimize distortion for each case. This article explores the selection process for near-linear features on the surface of the Earth and derives limits for the extent of a project that can be projected within specified distortion tolerances. It is then demonstrated that a multifaceted set of projections of the Earth may be used to extend this concept to the mapping of features such as highways and railways that are quasi-linear but do not exactly follow a standard geometrical line (a great circle or a small circle) on the surface of the Earth. A continuous, conformal coordinate system may be derived in such situations, extending to indefinite length and applicable over a swath of several kilometers width, but it cannot be extended to cover situations with extensive variations in height. Instead, the Snake Projection is analyzed, and it is shown that this can be used to develop continuous (non-zonal) projected coordinate systems for major engineering projects extending for hundreds of kilometers and having extensive height ranges. Examples are shown of the application to railway projects.     

IDENTIFICATION AND ANALYSIS OF NATURAL CHANNEL NETWORKS FROM DIGITAL ELEVATION MODELS

The identification and analysis of natural channel networks from digital elevation models are discussed from the point of view of their environmental applications. An interactive, graphical software package that implements some of the most widely used techniques for the automatic recognition of channel networks and for the computation of some useful geomorphologic indices and functions is presented.     

中国居民食物消费碳排放时空演变趋势及其驱动机制分析

食物生产是影响温室效应气体排放的重要方面,由于食物类型不同其碳排强度存在较大的差异,因此当前中国居民食物消费快速演替必然会对生态碳环境产生重要影响。论文利用洛伦兹曲线和基尼系数,研究1997—2016年中国城乡居民食物消费在生产过程中碳排放时空演变及其主要驱动因子。主要结论如下：① 在数量上,中国居民食物消费碳排放总量和人均量整体都呈现逐渐的增长刚性,但仍属于低碳消费模式;② 在结构上,各类型食物碳排放演变呈现显著的分异态势,整体上动物性食物碳排放增长速度快于植物性食物碳排放减少速度;③ 在城乡差异上,城镇和乡村居民食物消费碳排放在数量、结构和趋势上表现出显著的城乡二元属性,但近年城乡差异表现出减缓趋势;④ 在空间上,各地食物消费碳排放呈现较显著的民族性、地域性、集聚性和中心“塌陷”的特征,碳排放密度高值主要分布在直辖市和沿海经济相对发达地区;⑤ 在驱动力上,结构演替是全国居民食物消费碳排放总量整体变化的主要驱动因素,而人口数量是各地地域空间差异分布的主要驱动因子。     

机载测雨雷达两种反演方法的优化和综合利用的模拟研究

由于机载雷达的天线不能做得太大,如果想获得一个较好的束向分辨率,就不得不使用X波段的频率,而该波段的电磁波穿过雨区时会受到衰减,直接利用视反射率进行降水反演就会产生很大的误差,因此,衰减效应必须进行校正。两种降水反演方法——立体雷达方法和双束算法,在实际的反演中各有优缺点。立体雷达方法对边界条件要求高,而双束算法在雨团中心雨强大处会有较大的误差。引入权重因子后的立体雷达方法－双束算法(综合算法)能充分发挥两种方法的优点,从而更具适用性和准确性。     

Current-induced scour beneath initially elevated subsea pipelines

When a subsea pipeline is laid on an uneven seabed, certain sections may have an initial elevation with respect to the far-field seabed, e_o, and thus potentially affecting the on-bottom stability of the pipeline. This paper focuses on quantifying the effects of the upstream dimensionless seabed shear stress, θ_∞, and Reynolds number, Re, on: (1) the maximum dimensionless seabed shear stress beneath the pipe, θ_max, to be compared to the critical shear stress in order to determine whether scour would occur and progress towards an equilibrium state; and, (2) the dimensionless equilibrium scour depth beneath the pipe, S_eq/D. Using a 2-D Reynolds averaged Navier-Stokes (RANS) approach along with the k-ω Shear Stress Transport (SST) turbulence model, a parametric study involving 243 computational fluid dynamics (CFD) simulations was conducted. The simulation results were used to develop a closed-form equation for the prediction of θ_max. Subsequently, experimental measurements of S_eq/D have been compiled from published literature, to develop a new closed-form equation for the prediction of S_eq/D with a high correlation to the experimental data. In summary, we present two closed-form equations for the prediction of θ_max and S_eq/D for pipelines with an initial e_o/D, which are applicable for both clear-water and live-bed conditions. The effects of θ_∞ and Re have been included, albeit Re having a small influence as compared to the other parameters.     

雾霾的影响因子分析及其区域相关性研究

成都、重庆作为中国西南地区快速发展的城市,是雾霾频发的区域,研究其雾霾因子联系以及区域空间分布特征及其规律,对于城市发展、创建绿色城市具有重要意义。通过分析2016年12月~2017年2月的PM2.5等污染物浓度数据可知:①成渝地区雾霾产生的高峰期在1月,并以成都、重庆为中心向四周减弱;②在时间分布上,成都和重庆的雾霾污染物浓度呈协同变化趋势,说明两地雾霾在区域上相互影响,协同作用;③在气象因素中,两地的湿度和风速对雾霾集聚、转移和扩散具有重要作用;④两地的位置和地形特征是其雾霾具有相关性的重要因素,因此只有两地同时防治雾霾,才能达到最好的效果。根据不同地域特征,积极改进生产结构,采取相应措施,才是治理雾霾的根本所在。     

中国金红石矿床成矿地质条件及成矿远景

中国金红石矿床成因类型不一,控矿因素各异。通过对地层、构造(包括大地构造环境及褶皱)等控矿因素,区域变质作用、岩浆作用、热液作用、风化及沉积作用等成矿地质条件及矿床分布等规律的归纳与研究,提出了陕西石泉-镇坪、陕西丹凤-西峡、河南南召-方城、湖北枣阳-河南新县、湖北英山-潜山、江苏徐州-东海、山东栖霞-荣城、山西代县-河北涞水、四川会东9个金红石成矿远景区。并圈定出山东省荣城曲家、安徽凤阳板桥、湖北郧西土门—涧池、湖北七里坪大悟仙山、山西代县义成沟—羊廷寺、河北涞源司各庄、河北丰宁黄土梁等7个找矿靶区。     

基于贝叶斯正则化BP神经网络的DEM趋势面逼近

趋势面从宏观上揭示了研究对象的特性,在各领域发挥着重要作用。BP神经网络可以对复杂系统进行无限逼近,进而进行预测。建立了基于贝叶斯正则化BP神经网络的数字高程模型趋势面,与二次多项式建立的数字高程模型趋势面进行比较分析,证明了该方法的可行性和有效性。     

GA-BP网络模型在GPS似大地水准面精化中的应用

针对误差反向传播（BP）算法训练速度慢和易于陷入局部最小值的缺点,提出了利用遗传算法（GA）的全局寻优性,结合GA和BP的各自优点,分析和建立了进化神经网络（GA-BP）模型,并将该模型应用于似大地水准面模型精化。最后以南方某市E级GPS控制网高程数据为例,进行BP和GA-BP模型的对比实验,通过对内、外符合精度及MAPE（平均绝对误差百分比）指标分析,验证了该方法的可行性。