Detecting the dynamics of urban structure through spatial network analysis

Urban spatial structure in large cities is becoming ever more complex as populations grow in size, engage in more travel, and have increasing amounts of disposable income that enable them to live more diverse lifestyles. These trends have prominent and visible effects on urban activity, and cities are becoming more polycentric in their structure as new clusters and hotspots emerge and coalesce in a wider sea of urban development. Here, we apply recent methods in network science and their generalization to spatial analysis to identify the spatial structure of city hubs, centers, and borders, which are essential elements in understanding urban interactions. We use a ‘big’ data set for Singapore from the automatic smart card fare collection system, which is available for sample periods in 2010, 2011, and 2012 to show how the changing roles and influences of local areas in the overall spatial structure of urban movement can be efficiently monitored from daily transportation.

In essence, we first construct a weighted directed graph from these travel records. Each node in the graph denotes an urban area, edges denote the possibility of travel between any two areas, and the weight of edges denotes the volume of travel, which is the number of trips made. We then make use of (a) the graph properties to obtain an overall view of travel demand, (b) graph centralities for detecting urban centers and hubs, and (c) graph community structures for uncovering socioeconomic clusters defined as neighborhoods and their borders. Finally, results of this network analysis are projected back onto geographical space to reveal the spatial structure of urban movements. The revealed community structure shows a clear subdivision into different areas that separate the population’s activity space into smaller neighborhoods. The generated borders are different from existing administrative ones. By comparing the results from 3 years of data, we find that Singapore, even from such a short time series, is developing rapidly towards a polycentric urban form, where new subcenters and communities are emerging largely in line with the city’s master plan.

To summarize, our approach yields important insights into urban phenomena generated by human movements. It represents a quantitative approach to urban analysis, which explicitly identifies ongoing urban transformations.     

基于径向基函数神经网络的混合像元分解

遥感图像中普通存在着混合像元。对这部分像元进行分类（即混合像元分解）是遥感图像处理中的难点。基于主分量分析的混合像元分解 法是一种较为成熟的算法，但它存在着计算量大，适应性差等缺点。在深入研究混合像元分解原理的基础上，提出了用径向基函数神经网络拟合分解结果超平面，以实现混合像元分解的算法，实验结果证明：该算法的结果与基于主分量分析的混合像元分解算法结果相近（相关系数达到0．00），而计算量大大减少，具有较强的适应性。     

物联网与地理信息的集成应用研究

针对物联网技术及地理信息技术的特点,阐述了物联网与地理信息之间的联系.利用GIS特有的信息集成、可视化表达及强大的空间分析等技术特点,把地理信息技术与物联网技术结合,提出了由感知层、数据层、网络层、服务层、应用层五层结构组成的基于GIS的物联网信息系统的初步设计思路.     

Effects of hillslope dynamics and network geometry on the scaling properties of the hydrologic response

This paper investigates the specific contributions of river network geomorphology, hillslope flow dynamics and channel routing to the scaling behavior of the hydrologic response as function of drainage area. Scaling relationships emerged from the observations of geomorphological and hydrological data and were reproduced in previous works through mathematical models, for both idealized self-similar networks and natural basins. Recent literature highlighted that scale invariance of hydrological quantities depends not only on the metrics of the drainage catchment but also on effective flow routing. In this study we employ a geomorphological width function scheme to test the simple scaling hypothesis adopting more realistic dynamic conditions than in previous approaches, specifically taking into account the role of hillslopes. The analysis is based on the derivation of the characteristic distributions of path lengths and travel times, inferred from DEM processing and measurements of rainfall and runoff data. The study area is located in the Tiber River region (central Italy).Results indicate that, while scaling properties clearly emerge when the hydrologic response is defined on the basis of the sole geomorphology, scale invariance is broken when less idealized flow dynamics are taken into account. Lack of scaling appears in particular as a consequence of the catchment to catchment variability of hillslope velocities.     

人工神经网络泛化性能改进

泛化能力是人工神经网络的重要特性.总结了增强神经网络泛化能力的几种方法,分析了各自的优缺点,并使用Matlab中提供的函数进行非线性函数的逼近,通过仿真与原函数拟合验证了泛化能力的提高．     

基于M16C单片机的温度测量计

研制了一种模块化设计的瑞萨系列单片机实验系统.系统采用集成度高的温度传感器,运用模块化设计完成了基于16位单片机(M3062LFGPGP)的数字式温度计设计,同时在处理温度数据时运用一种新算法,温度精度通过主控模块软件和温度硬件校正处理得到提高.根据传统设计原理,结合瑞萨单片机M3062LFGPGP特点,系统将A/D转换模块和万年历模块集中于主控模块,实现了用软件编程替代A/D模块、万年历硬件模块的新型设计.设计需要模块仅为主控模块、温度模块和显示模块,能够显示当前环境温度以及年、月、日、星期、时、分、秒.此设计可以大大降低系统功耗,减少硬件设计错误,提高系统的稳定性,计时准确,显示清晰,具有较高的应用价值.     

新一代温室气体观测卫星（GOSAT、OCO）传感器设置

为了深入了解国际上新一代CO2等温室气体的空间观测计划及进展情况,以日本和美国于2009年最新发射的温室气体测卫星＂呼吸＂号（GOSAT）、＂轨道碳观测者＂（OCO）为例,分析了观测计划的开展情况,对卫星上搭载的传感器设置做了分析,对新一代温室气体专用观测卫星与传统卫星上的传感器做了对比,得出了新一代的传感器在设置上的特点：波段设置在近红外太阳辐射以获得边界层的温室气体含量,具有高的光谱分辨率以保证观测精度,多种观测方式结合以获得最大的信息量。并提出了我国应对温室气体观测资料缺乏的对策,发展专用观测卫星或传感器,以提供我们自己的高精度的全球和区域温室气体观测数据。     

Scalable framework of intelligent RFI flagging for large-scale HI survey data from FAST

Radio frequency interference (RFI) identification is a key step in radio data processing. In order to efficiently process huge volumes of data produced by modern large radio telescopes, such as the Five-hundred-meter Aperture Spherical radio Telescope (FAST), exceptional balance between accuracy and performance (throughput) is required for RFI flagging algorithms. RFI-Net is a single-process RFI identification package based on deep learning technique, and has achieved a higher flagging accuracy than the classical SumThreshold method. In this paper, we present a scalable RFI flagging toolkit, which can drive parallel workflows on multi-CPU and multi-GPU clusters, with RFI-Net as its core detector. It can automatically schedule the workload and aggregate itself after errors according to the running environment. Moreover, its main components are all pluggable, and can be easily customized according to requirements. The experiments with real data of FAST showed that using eight parallel workflows, the toolkit can process sky survey data at a speed of 66.79 GB/h, which means quasi-real-time RFI flagging can be achieved considering the data rate of FAST extragalactic spectral line observations.     

预应力锚索锚固段钢绞线应力分布研究

使用磁通量传感器对拉力集中型预应力锚索锚固段钢绞线应力分布进行研究,分析了预应力锚索锚固段钢绞线应力传递和分布的规律;结合长期监测,分析了锚固段钢绞线应力损失情况以及张拉锁定前后锚固段钢绞线应力分布调整的规律。研究应用表明,磁通量传感器技术可作为锚索基本试验等各种试验中钢绞线应力分布检测的一种有效的方法。     

基于社团的移动容迟网络源路由算法

针对移动容迟网络中的数据共享服务,提出基于社团的源路由算法(Social-based Source Routing,SSR)。将移动容迟网络中的数据共享过程分为摘要消息广播、兴趣消息回传与内容数据转发。利用分布式社团检测算法区别各节点的多社团结构,通过摘要消息广播构建节点间的社团路径(Community Path),并将共享数据沿社团路径进行基于单消息副本的转发。仿真结果表明算法在一定条件下能够达到与多副本转发算法类似的消息传输成功率,同时显著降低消息传输代价。