基于相似数据聚合与变K值KNN的短时交通流量预测

短时交通流量预测是交通控制和诱导涉及的关键技术问题,由于短时交通流量存在不确定性和时变性,其预测难度较大,是相关研究领域与工程实践中亟待解决的难题。为提高短时交通流量预测的准确性,本文设计与实现了基于相似数据聚合和变K值KNN(KNN-SDA)的短时交通流量预测算法。该算法首先采用互信息法在经过预处理的交通流量数据集提取交通流量序列最佳延迟时间信息,生成状态向量,并构建交通流量历史数据库;然后以本文所提出的相似数据聚合方法完成历史数据的聚合与清洗得到训练数据集;最后通过交叉验证确定每个时刻的最优K近邻数,完成算法实现。实验结果表明,本文提出的变K值KNN-SDA算法在保证执行效率的同时能明显提高短时交通流量的预测精度。     

Are all patterns created equal? Cooperation is more likely in spatially simple habitats

Cooperative behaviours, such as aggregation with neighbouring conspecifics, can enhance resilience in habitats where risks (i.e. predation, physical disturbances) are high, exerting positive feedback loops to maintain a healthy population. At the same time, cooperation behaviours can involve some extra energy expenditures and increasing resource competition. For sessile reefs, like mussels, simulation models predict increased cooperation under increasing levels of environmental stress. Predation risk is viewed as a behaviour‐modifying stressor, but its role on cooperation mechanisms, such as likelihood of reciprocity, has not yet been empirically tested. This study harnesses this framework to understand how cooperation changes under different perceived levels of predation risk, using mussel beds as model of a complex “self‐organised” system. Hence, we assessed the context dependency of cooperation response in different “landscapes of fear,” created by changes in predator cues, substratum availability and body size. Our experiments demonstrated that i) cooperation in a mussel bed system increases when predator cues are present, but that this relationship was found to be both, ii) strongly context‐dependent, particularly upon substratum availability and iii) size‐dependent. That is, while cooperation is in general greater for larger individuals, the response to risk results in greater cooperation when alternative attachment substratum is absent, meaning that simpler landscapes may be perceived as riskier. The context dependency of structural complexity is also an essential finding to consider in a changing world where habitats are losing complexity and cooperative strategies should be maximised.     

制图综合中建筑物多边形的合并与化简

讨论了数字环境下顾及建筑物矩形几何特征的多边形自动综合算法,针对多边形之间的拓扑邻近与视觉邻近两种空间关系,提出了基于矢量和基于栅格的两种建筑物多边形合并方法。关于建筑物形状的化简,本文提出了矩形差分方法,并在此基础上建立了建筑物多边形化简的层次化途径。     

等高线的空间关系规则和渐进式图形简化方法

详细讨论了等高线表达地形的规则和以此为基础自动建立等高线关系的方法,对地形特征点、线的提取改进了已有较成熟的方法,建立了一套实用的等高线图形简化的渐进式方法,并对其特殊情况的处理提 出了具体的算法。这种方法把不同比例尺跨度的等高线图形综合融为一体,易于实现,等高线图形简化时的等高线相交可在综合过程中控制。     

Cross-Level Inference in Political Science

Political Science research encounters inferences across levels of analysis; however, they are fraught with challenges. After introducing voting examples of aggregation bias, problems posed by aggregation bias are summarized more generally. Subsequently, the article reviews the major methodological approaches to overcome aggregation bias and to solve the ecological inference (disaggregation) problem. The article highlights the possibility that aggregation bias may lead governments to accept (or reject) international climate agreements when negotiating as blocs of countries as compared to the distribution of the preferences of all countries involved in the negotiations.     

面向流媒体传输的空间数据变化累积模型

将流媒体技术引入到以矢量数据为主的网络GIS传输中,其关键是建立逐级精细化的多层次表达模型,并在服务器端数据预组织上建立线性索引结构.本研究针对该目标建立一种矢量数据多重表达的新模型,即变化累积模型,认为尺度空间中从粗到细的数据表达过程是一些"变化"的累积,"变化"表现为两个连续尺度表达间的差别,通过"变化"元素的逐渐增加或减少来实现渐进传输.基于该模型对面状数据渐进式传输进行研究,将多边形分级剖分为一系列凸壳或外接矩形的基础上建立变化累积模型,通过剖分元素的组合完成渐进式流媒体传输.

Abstract：

The streaming transfer method which makes data immediately available application during the progress of transfer becomes a welcome transfer way over Web. The progressive transmission of map data provides the users witha self-adaptive strategy to access remote data. It not only speeds up the Web transfer but also offers an efficient navigation guide in information acquisition. The key technology in this transmission is the efficient multiple representation of spatial data and pre-organization on server site. This paper aims at vector data offering a new multiple representation model called changes accumulation model, which considers the spatial representation from one scale to another scale as an accumulation of set of changes. The difference between two consecutive representations is recorded in a linear index and through gradually addition or subtraction realizes the progressive transmission. As an example, the progressive transmission of area feature based on changes accumulation model is investigated in the paper. The model is built by hierarchical decomposition of polygon into series of convex hulls or bounding rectangles and through component of the decomposed elements to accomplish the progressive transmission.     

基于城市产业结构特征的高铁生产性服务业集散效应——以长三角城市群为例

高铁对生产性服务业的空间集散效应深受城市特征的影响。根据城市产业发展特征将城市分为二产主导型城市和三产主导型城市,借助双重差分模型探讨高铁对不同产业结构的城市生产性服务业及其细分行业的空间效应差异。结果表明,高铁对生产性服务业及细分行业的集散效应具有显著时滞性和异质性,城市产业结构对其效应产生了显著影响。① 基于生产性服务业整体的研究表明,在静态上,高铁促进二产主导型城市产业空间分散、三产主导型城市产业空间集聚;在动态上,高铁对二产主导型城市的生产性服务业空间集散具有先促进集聚后促进分散的特性,对三产主导型城市的产业效应始终表现为促进集聚。② 对于运输服务业、租赁与商服业而言,无论是静态效应还是动态效应,城市产业结构对其影响均较小,在静态上高铁促进二产主导型城市和三产主导型城市的产业在空间分散布局;在动态上高铁效应总体表现为“不显著-促进分散”的二阶段性。③ 对于金融服务业和研发设计业而言,高铁的静态效应在作用方向上具有一致性,均表现为促进产业集聚,其动态效应受到城市产业结构的影响表现出显著阶段性差异。④ 对于信息服务业而言,高铁对其空间集散的静态效应表现为对二产主导型城市的产业影响不显著,促进三产主导型城市的信息服务业在空间集聚,且其动态效应与静态效应相一致。     

中国县域可达性区域划分及其与人口分布的关系

可达性空间划分是交通地理学研究的重要领域。分析其与人口布局的关系,对当下中国的发展具有重要的理论与实践意义。借助GIS 软件平台,运用栅格成本加权距离算法,计算全国县 (市、区) 域单元的空间可达性并进行区域划分,进而探讨中国大区域交通可达性与人口分布的关系,揭示不同交通条件和不同区域类型中人口迁移与人口流动内在机制,为区域规划、交通规划和中国人口配置提供科学依据。研究表明,①栅格成本加权距离算法可以实现空间可达性概念从定性描述转向定量计算的精确转换;②中国县域可达性以低值为主,具有圈层状和"反自然梯度"空间结构,"黑河-腾冲"线为明显分界线;③中国大陆分为3大区:可达性高值区、可达性中值区、可达性低值区。可达性中值区又可分为东南诸省亚区、陕甘宁晋亚区、吉林亚区;可达性低值区分为云川渝亚区、蒙黑亚区、新疆亚区、青藏亚区;④国家区域范围内县域可达性与人口密度具有明显的相关性,2.5小时圈是目前中国可达性与人口集聚产生作用的最远距离;中国县域可达性越高的区域,可达性对人口集聚产生作用圈就越大,而可达性值与人口数量的相关性越小,说明交通条件对落后地区人口集聚所起到的作用远远大于发达地区。     

冀东长城式金矿构造地质特征及构造演化与金矿化的关系

长城式金矿是产在中晚元古宇碳酸岩中的一种新类型金矿,本文从金矿床的构造地质特征及其演化的角度,阐述了长城式金矿构造与金矿化的独特关系。     

边坡破坏过程离心模型试验的应力位移场研究

提出了一种确定离心模型试验过程中边坡的应力和位移场及其变化过程的方法。基于物理测量与数值模拟相结合的思路,首先采用离心场非接触位移测量技术测量出边坡的位移场;然后通过数值模拟和反演分析等途径算出边坡的应力场。该方法已成功用于多个边坡离心模型试验中的应力位移场的确定,表明了该方法可以较好地得出离心模型试验过程中边坡的应力和位移场。边坡离心模型试验结果表明,边坡的破坏过程与应变局部化和应力集中的过程有着密切联系。