Spatial aggregations for spatial based decision making in spatial data warehouses

For spatial based decision making such as choice of best place to construct a new department store, spatial data warehousing system is required more and more previous spatial data warehousing systems; however, provided decision making of non-spatial data on a map and so those cannot support enough spatial based decision making. The spatial aggregations are proposed for spatial based decision making in spatial data warehouses. The meaning of aggregation operators for applying spatial data was modified and new spatial aggregations were defined. These aggregations can support hierarchical concept of spatial measure. Using these aggregations, the spatial analysis classified by non-spatial data is provided. In case study, how to use these aggregations and how to support spatial based decision making are shown.

Fast Web - Based Data Transmission

Since web based GIS processes large size spatial geographic information on internet, we should try to improve the efficiency of spatial data query processing and transmission. This paper presents two efficient methods for this purpose: division transmission and progressive transmission methods. In division transmission method, a map can be divided into several parts, called “tiles”, and only tiles can be transmitted at the request of a client. In progressive transmission method, a map can be split into several phase views based on the significance of vertices, and a server produces a target object and then transmits it progressively when this spatial object is requested from a client. In order to achieve these methods, the algorithms, “tile division”, “priority order estimation” and the strategies for data transmission are proposed in this paper, respectively. Compared with such traditional methods as “map total transmission” and “layer transmission”, the web based GIS data transmission, proposed in this paper, is advantageous in the increase of the data transmission efficiency by a great margin.

地理信息技术赋能国土空间规划实施监督和体检评估预警

针对地理信息技术赋能国土空间规划实施监督和体检评估预警进行了研究论述,围绕强化规划权威和规划实施监督管理工作需要,从汇聚规划成果、强化实施监督、赋能体检评估3个角度进行深入研究,引入了新技术赋能国土空间规划实施监督与动态感知,完善了国土空间规划体检评估预警体系。

“一网统管”城市运行一张图设计与实现

基于智慧城市和数字政府建设对城市精细化管理提出的更高要求,为夯实“一网统管”数字底座基础,结合常州实际,设计了城市运行一张图的总体框架,建立全市统一的空间数据底座,形成资源展示、数据共享、场景配置、二次开发、后台管理等能力。通过应用实践,有效加快了城市空间、城市部件、城市运行动态的数字化进程,促进数据共享、治理转型,支撑政府部门各类智能化应用,推动数字政府转型和城市建设智慧化升级。

基于改进DBSCAN和距离共识评估的分段点云去噪方法

针对点云数据中噪声点的剔除问题,提出了一种基于改进DBSCAN(density-based spatial clustering of applications with noise)算法的多尺度点云去噪方法。应用统计滤波对孤立离群点进行预筛选,去除点云中的大尺度噪声;对DBSCAN算法进行优化,减少算法时间复杂度和实现参数的自适应调整,以此将点云分为正常簇、疑似簇及异常簇,并立即去除异常簇;利用距离共识评估法对疑似簇进行精细判定,通过计算疑似点与其最近的正常点拟合表面之间的距离,判定其是否为异常,有效保持了数据的关键特征和模型敏感度。利用该方法对两个船体分段点云进行去噪,并与其他去噪算法进行对比,结果表明,该方法在去噪效率和特征保持方面具有优势,精确地保留了点云数据的几何特性。

基于机器学习空间聚类的出租车停靠站点布局规划

针对出租车随意停靠给城市交通带来的负面影响,为规范出租车营运秩序、改善出租车营运环境和居民乘车条件,提出一种将出租车出行空间信息与机器学习算法相结合的出租车停靠站点布局规划方法。首先利用出租车GPS轨迹数据提取出租车出行起点,然后采用HDBSCAN聚类算法对起点进行空间密度聚类,形成聚类簇后以其中心点作为出租车停靠站点布局的备选点。最后,为验证所提方法的可行性和有效性,选取重庆市中心城区一土地利用类型丰富、人口密度高的典型区域进行案例分析。结果显示,107个备选点主要分布于商业中心区和居住集中区,与出租车出行高需求区域的空间分布基本吻合;布局的出租车停靠站点在300 m范围内的覆盖率达到76.0%,未覆盖区域主要为城市绿地和水体。研究表明,机器学习算法可实现出租车停靠站点的高效布局规划,但在规划和实施阶段,停靠站点的设置还应结合邻近区域的建成环境特点综合考虑。

面向自然资源要素三维管理的关键技术研究与应用

针对省级自然资源要素三维管理中时空数据处理传统方法存在数据处理耗时长、工序复杂、三维浏览效率低下等问题,开展自然资源要素三维管理关键技术研究。基于自然资源相关多源异构数据,研究基于镶嵌数据集的TB级海量影像管理技术、基于CRF压缩存储技术的影像快速切片技术、支持并行的三维场景实时渲染技术等技术方法,提高了自然资源要素三维时空数据处理、可视化效率;并研发自然资源要素三维管理系统,实现自然资源要素数据三维空间基底,为自然资源管理相关业务应用提供支撑。

“一带一路”沿线国家工程建设信息系统的建设

海外工程建设面临诸多地质风险,本文提出了一种以空间数据库和地理信息系统(GIS)技术为支撑的信息系统建设方式,将已收集的“一带一路”沿线国家数百项工程、上千个地质钻孔、6种特殊岩土及多种专题地质图件等数据进行统一入库管理,并集成了海外工程建设标准规范、施工工艺等非结构化数据,实现了工程建设多源异构数据源的一体化管理。该信息系统基于缓冲区分析和叠加分析等GIS空间分析技术,可快速获取在建工程项目及周边场地的综合地质信息,为工程建设全过程提供数据支撑,提高了数据的应用效率。通过将各类数量巨大、来源分散、格式多样的工程建设数据进行统一采集、存储、关联分析及智能应用,打破了数据孤岛,不仅可为单一工程建设提供服务,还可为大区域范围内的数据挖掘、知识发现及智能分析等应用打下基础。

城镇化推进模式与我国农业低碳全要素生产率——来自双源夜间灯光证据

本文利用卫星监测的数据构造夜间灯光复合指数表征城镇化水平,运用Superefficiency Ray Slacks-Based Measure(Super-RSBM)模型和Global Malmquist-Luenberger(GML)指数测算2000—2021年我国农业低碳全要素生产率(TFP),实证检验城镇化对我国农业低碳TFP的影响及其作用机制,并考察紧凑集约型和规模扩张型两种城镇化推进模式对农业低碳TFP的异质性影响。研究发现,从全国来看,城镇化推进与农业低碳TFP之间具有显著的U型关系,且邻近地区农业低碳TFP的提升对本地区产生示范效应;分区域来看,这种U型关系主要体现在农业适度发展区,而农业优化发展区的城镇化与农业低碳TFP之间呈现显著的正向线性关系,表明农业优化发展区应发挥“领头羊”作用,带动适度发展区早日跨越U型曲线的拐点,实现城镇化带动农业绿色发展;紧凑集约型的城镇化深度推进模式能够显著提升农业低碳TFP,而规模扩张型的城镇化广度推进模式降低了农业低碳TFP;农业低碳技术进步、农村劳动力转移、规模效应、农业产业链延伸和农村居民可支配收入增加是城镇化影响农业低碳TFP的主要途径。

Over-sampling algorithm for imbalanced data classification

For imbalanced datasets, the focus of classification is to identify samples of the minority class. The performance of current data mining algorithms is not good enough for processing imbalanced datasets. The synthetic minority over-sampling technique(SMOTE) is specifically designed for learning from imbalanced datasets, generating synthetic minority class examples by interpolating between minority class examples nearby. However, the SMOTE encounters the overgeneralization problem. The densitybased spatial clustering of applications with noise(DBSCAN) is not rigorous when dealing with the samples near the borderline.We optimize the DBSCAN algorithm for this problem to make clustering more reasonable. This paper integrates the optimized DBSCAN and SMOTE, and proposes a density-based synthetic minority over-sampling technique(DSMOTE). First, the optimized DBSCAN is used to divide the samples of the minority class into three groups, including core samples, borderline samples and noise samples, and then the noise samples of minority class is removed to synthesize more effective samples. In order to make full use of the information of core samples and borderline samples,different strategies are used to over-sample core samples and borderline samples. Experiments show that DSMOTE can achieve better results compared with SMOTE and Borderline-SMOTE in terms of precision, recall and F-value.

Using ALS raster data in forest planning

Raster type of forest inventory data with site and growing stock variables interpreted for small squareshaped grid cells are increasingly available for forest planning.In Finland,there are two sources of this type of lattice data:the multisource national forest inventory and the inventory that is based on airborne laser scanning(ALS).In both cases,stand variables are interpreted for 16 m×16 m cells.Both data sources cover all private forests of Finland and are freely available for forest planning.This study analyzed different ways to use the ALS raster data in forest planning.The analyses were conducted for a grid of 375×375 cells(140,625 cells,of which 97,893 were productive forest).The basic alternatives were to use the cells as calculation units throughout the planning process,or aggregate the cells into segments before planning calculations.The use of cells made it necessary to use spatial optimization to aggregate cuttings and other treatments into blocks that were large enough for the practical implementation of the plan.In addition,allowing premature cuttings in a part of the cells was a prerequisite for compact treatment areas.The use of segments led to 5–9%higher growth predictions than calculations based on cells.In addition,the areas of the most common fertility classes were overestimated and the areas of rare site classes were underestimated when segments were used.The shape of the treatment blocks was more irregular in cell-based planning.Using cells as calculation units instead of segments led to 20 times longer computing time of the whole planning process than the use of segments when the number of grid cells was approximately 100,000.

Soil polygon disaggregation through similarity-based prediction with legacy pedons

Conventional soil maps generally contain one or more soil types within a single soil polygon.But their geographic locations within the polygon are not specified.This restricts current applications of the maps in site-specific agricultural management and environmental modelling.We examined the utility of legacy pedon data for disaggregating soil polygons and the effectiveness of similarity-based prediction for making use of the under-or over-sampled legacy pedon data for the disaggregation.The method consisted of three steps.First,environmental similarities between the pedon sites and each location were computed based on soil formative environmental factors.Second,according to soil types of the pedon sites,the similarities were aggregated to derive similarity distribution for each soil type.Third,a hardening process was performed on the maps to allocate candidate soil types within the polygons.The study was conducted at the soil subgroup level in a semi-arid area situated in Manitoba,Canada.Based on 186 independent pedon sites,the evaluation of the disaggregated map of soil subgroups showed an overall accuracy of 67% and a Kappa statistic of 0.62.The map represented a better spatial pattern of soil subgroups in both detail and accuracy compared to a dominant soil subgroup map,which was commonly used in practice.Incorrect predictions mainly occurred in the agricultural plain area and the soil subgroups that are very similar in taxonomy,indicating that new environmental covariates need to be developed.We concluded that the combination of legacy pedon data with similarity-based prediction is an effective solution for soil polygon disaggregation.

城市“空间数字底座”发展趋势及上海的建设思路

“空间数字底座”有效促进空间信息研发团队集聚,构建更大范围区域一体化空间信息创新链产业链,支撑孵化具有我国自主知识产权的空间信息服务软硬件体系,产生一批具有技术实力的空间信息上下游企业。“空间数字底座”建设为上海巩固提升城市能级和核心竞争力、构筑上海未来战略新优势提供坚实支撑,将增强上海空间信息产业服务能级,进一步带动空间信息、航空航天、智能驾驶、节能环保、数字物流等产业发展。

复杂工况下综合传动装置状态监测数据异常检测方法

针对综合传动装置运行过程中,工况变化及装置故障状态引起的数据异常、难以有效区分这一问题,提出了一种适用于复杂工况下综合传动装置状态监测数据异常检测的方法。首先,采用基于密度的聚类方法(DBSCAN)对状态监测数据进行了关联变量聚类,以排除非关联数据对数据重构准确度的干扰;然后,利用深度降噪自编码网络构建了状态监测数据重构模型,获取了对异常数据敏感的偏差特征;最后,利用支持向量数据描述(SVDD)算法构建了正常状态监测数据偏差特征的超球体,完成了复杂工况下对综合传动装置状态监测数据异常的检测;为了验证该方法对综合传动装置状态监测数据异常检测的有效性,以某型综合传动装置为研究对象,在多组综合传动装置漏油实验数据上进行异常检测验证分析。实验结果表明:该方法实现了在综合传动装置不同程度漏油故障条件下对状态监测数据异常进行检测的目的,且其准确度整体高于92%。研究结果表明:该方法可以有效检测出综合传动装置早期异常运行状态,为综合传动装置健康管理与劣化评估奠定基础。

GIS技术在国土空间规划中的探索

在国土空间规划领域,为了全面落实国家和地方对信息基础设施建设的要求,同时兼顾国家“放管服”改革需求,应充分运用地理信息(GIS)技术,以国土空间管控数字化、地图可视化为目标,实现“多规合一”及全要素的管控。采用GIS技术,通过国土空间规划全周期二维、三维一体化GIS数据底座和平台建设,让自然资源规划、管理、实施监督、应用全流程更加系统化,推动国土空间规划数字化转型升级,助力国土空间高质量发展。

基于聚类算法的电网告警数据分析与处理模型

提出了一种基于时间戳和关键字的聚类算法来解决告警数据种类繁多且难以提取关键信息的问题。首先,对告警数据中的最新发生时间进行K-Means聚类;其次,基于告警数据开始时间进行K-Means二次聚类;再次,使用具有噪声的基于密度的聚类算法(Density-Based Spatial Clustering of Application with Noise,DBSCAN)对每列关键字进行聚类;最后,对结果进行了整合,并给出了关联性描述结果。实验结果表明,通过上述聚类算法构建的告警数据分析与处理模型的平均压缩率为79.28%,平均准确率达到93.41%,能够有效提高对现有告警数据的具象化描述能力,降低告警数据理解的复杂度。

基于神经网络的行驶时长预测

定位设备的普及产生了海量的车辆行驶数据,使得利用历史数据预测车辆行驶时长成为可能.车辆行驶数据由两部分组成:车辆行驶经过的路段序列信息和出发时段;路径总长度等外部信息.如何提取路段序列特征,以及如何将序列特征与外部特征有效地融合,成为预测行驶时间的关键问题.为解决以上问题,提出了一个基于Transformer的行驶时间预测模型,模型由路段序列处理模块和特征融合模块两部分组成.首先,路段序列处理模块使用自注意力机制处理路段序列,提取路段序列特征.该模型不但可以充分考虑各条路段与其他路段间道路速度的时空关联性,同时可保证数据并行输入模型,避免了使用循环神经网络时数据顺序输入导致的效率低下.其次,特征融合模块将路段序列特征与出发时段等外部信息相融合,最终获得预测的行驶时长.在此基础上,统计路口连接的路段数作为路段的上/下游路口特征,与路段特征结合输入模型,进一步提升了行驶时长的预测精度.在真实的数据集上与主流预测模型进行的对比实验表明,该模型在预测精度以及训练速度上均有提升,体现了所提模型的有效性.

自然灾害视角下区域韧性评价方法及影响因素

为评估不同地区面对自然灾害冲击的区域韧性及影响因素,以自然灾害冲击地区所表现出的脆弱性和适应性,综合反映区域韧性,构建一种理解和量化区域韧性的框架。运用基于松弛变量的数据包络分析(SBM-DEA)方法,测算2011—2020年我国27个省、自治区、直辖市的区域韧性水平;并结合空间杜宾模型,综合考察社会、经济、环境等多维影响因素与区域韧性的空间效应关系。研究结果表明:自然灾害视角下全国区域韧性水平呈现北低南高的整体格局,地区间差异较大,年份差异相对较小,但变化幅度和频率存在逐渐扩大的趋势,区域韧性表现并不与社会经济发展水平强相关;空间计量回归中,教育程度、医疗资源、社会保障、产业结构等多种影响因素与区域韧性形成显著地空间效应关系。

基于DBSCAN算法的海战场环境下目标分群

针对海战场环境下态势评估中目标数量多、类型复杂多样的问题,首先引入数据聚类对态势评估的目标分群环节进行聚类分群,提出基于DBSCAN(density-based spatial clustering of applications with noise)算法的密度聚类,可聚类任意形状的数据簇,遍历性好,能够对战场环境下目标进行全面合理的分群;然后,给出了算法计算的基本步骤,并利用算例对已知战场态势的目标群进行正确性验证;最后,将该算法与基于划分的K-means算法、基于层次的AGNES(AGglomerative NESting)算法进行了对比分析,证明了该算法的有效性和合理性。

基于DBSCAN聚类的热能发电大数据异常检测模型

为了解决热能发电大数据异常检测时,存在误报率高、检测率低和漏报率高的问题,提出了基于DBSCAN聚类的热能发电大数据异常检测模型。首先通过时间序列模型对原始发电大数据进行异常值修正,然后将修正后的数据归一化处理,最后基于Spark Streaming设计Streaming DBSCAN算法,结合历史数据和相似发电厂数据的聚类特征,完成热能发电异常数据的检测。实验结果表明,所提方法可以有效降低误报率和漏报率、提高检测率并准确地获取异常值。