New Rural Community Construction or Retention Development:A Comparative Analysis of Rural Settlement Transition Mechanism in Plain Agriculture Area of China Based on Actor Network Theory

It is an important way to realize rural revitalization and sustainable development to guide rural settlement transition(RST)in an appropriate way.This paper uses actor network theory(ANT)to construct a theoretical framework for the study of RST.Taking two typical villages with different transition paths in rural areas of North China Plain as examples,this paper reveals the mechanism of RST and makes a comparative analysis.The results show that:1)after identifying problems and obligatory passage point,key actors recruit heterogeneous actors into the actor network by entrusting them with common interests,and realize RST under the system operation.2)Rural settlements under different transition paths have similarities in the problems to be solved,collective actions and policy factors,but there are differences in the transition process,mechanism and effect.The actor network and mechanism of RST through the path of new rural community construction are more complex and the transition effect is more thorough.In contrast,the degree of RST of retention development path is limited if there is no resource and location advantage.3)Based on the applicable conditions of different paths,this paper designs a logical framework of‘Situation-Structure-Behavior-Result’to scientifically guide the identification of RST paths under the background of rural revitalization.

Identifying Restructuring Types of Rural Settlement Using Social Network Analysis:A Case Study of Ezhou City in Hubei Province of China

Social interaction has become one of the key factors affecting the spatial reconstruction of rural settlements(SRRS).However,most studies ignored the multi-scale impact of social networks on the identification of restructuring types of rural settlements.This paper,taking Ezhou City of Hubei Province,China as the case study area,developed a potential inter-settlement network through considering settlements as nodes,and inter-settlement interactions induced by the spatial disparity of public facilities as edges,divided towns in Ezhou City into three zones based on community structure at the town level,and then identified four types of rural settlements in light of the characteristics of cluster patterns and centrality at the patch level.The results show that the inter-settlement network in Ezhou City presents apparent disparities in terms of community structure,cluster patterns and centrality.In community analysis,high inter-community and intra-community interactions are concentrated in well-developed areas in the north and east,while weak interactions between communities occur in the southern areas dominated by traditional agricultural production.Accordingly,three zones are divided such as the urban-leading zone,urban-rural integration zone and rural-leading zone.For the network centrality and cluster patterns,high-level rural settlements are mainly distributed in the urban-leading zone,followed by the urban-rural integration zone and the rural-leading zone.Moreover,the lump cluster pattern is observed in each zone,but the chain pattern and dispersed pattern largely occur in the rural-leading zone.At same time,four types of rural settlements are identified,namely urbanized settlements,central settlements,grassroots settlements and relocated settlements.The corresponding plans are discussed in different zones regarding urbanization,integration and characteristics to provide meaningful insights for policymakers to guide SRRS.This study would contribute to our understanding of the impact of social network involved in daily life on rural settlement reconstruction,and expect to provide theoretical and methodological support for rural sustainable development in practice.

Human settlement's social network analysis based on spatial trend surface model:a case study of Chongqing Municipality

Integrated with GIS and remote sensing(RS) technology,a systematic analysis and its methodology for human-settlements social environment has been introduced.This methodology has been called spatial trend field model(STFM).STFM's application history in the field of human-settlements social environment has been discussed at first.Then,some index data models have been created through STFM,which include population density trend field,human activity strength trend field,city-town spatial density trend field,urbanization ratio trend field,road density trend field,GDP spatial density trend field and PER-GDP spatial density trend field.With all above-mentioned indexes as input data,through Iterative Self-Organizing Data Analysis Techniques Algorithm(ISODATA),this paper makes a verification study of Chongqing municipality.The result of the case study confirms that STFM methodology is credible and has high efficiency for regional human-settlements study.

基于GIS-Network Analyst的重庆城市公园绿地可达性分析

依据大型城市公园绿地斑块、道路网和居民聚居点分布数据,结合航片解译的绿地覆盖数据,利用Arc GIS9.3的Network Analyst工具,分析了重庆市主城中心城区在30min步行距离内街道居民到达城市公园绿地斑块的可达性以及城市公园绿地斑块的服务指数.结果表明:重庆市主城中心城区规划范围分别有78.8%,55.8%和34.6%的街道居民在步行30min,20min和10min距离内能够到达大型城市公园绿地斑块;从街道居民点的行政归属看,以渝中区、渝北区、江北区和南岸区的居民可达性最好,大渡口区和沙坪坝区的居民区可达性较差.

Predicting and validating the load-settlement behavior of large-scale geosynthetic-reinforced soil abutments using hybrid intelligent modeling

Settlement prediction of geosynthetic-reinforced soil(GRS)abutments under service loading conditions is an arduous and challenging task for practicing geotechnical/civil engineers.Hence,in this paper,a novel hybrid artificial intelligence(AI)-based model was developed by the combination of artificial neural network(ANN)and Harris hawks’optimisation(HHO),that is,ANN-HHO,to predict the settlement of the GRS abutments.Five other robust intelligent models such as support vector regression(SVR),Gaussian process regression(GPR),relevance vector machine(RVM),sequential minimal optimisation regression(SMOR),and least-median square regression(LMSR)were constructed and compared to the ANN-HHO model.The predictive strength,relalibility and robustness of the model were evaluated based on rigorous statistical testing,ranking criteria,multi-criteria approach,uncertainity analysis and sensitivity analysis(SA).Moreover,the predictive veracity of the model was also substantiated against several large-scale independent experimental studies on GRS abutments reported in the scientific literature.The acquired findings demonstrated that the ANN-HHO model predicted the settlement of GRS abutments with reasonable accuracy and yielded superior performance in comparison to counterpart models.Therefore,it becomes one of predictive tools employed by geotechnical/civil engineers in preliminary decision-making when investigating the in-service performance of GRS abutments.Finally,the model has been converted into a simple mathematical formulation for easy hand calculations,and it is proved cost-effective and less time-consuming in comparison to experimental tests and numerical simulations.

Analysis and Prediction of Foundation Settlement of High-Rise Buildings under Complex Geological Conditions

Based on an example of a project in Tangshan, the high-rise buildings are built in karst area and mined out affected area which is treated by high pressure grouting, and foundation is adopted the form of pile raft foundation. By long-term measured settlement of high-rise buildings, It is found that foundation settlement is linear increase with the increase of load before the building is roof-sealed, and the settlement increases slowly after the building is roof-sealed, and the curve tends to converge, and the foundation consolidation is completed. The settlement of the foundation is about 80% - 84% of the total settlement before the building is roof-sealed.Three layer BP neural network model is used to predict the settlement in the karst area and mined affected area.Compared with the measured data, the relative difference of the prediction is 0.91% - 2.08% in the karst area, and is 0.95% - 2.11% in mined affected area. The prediction results of high precision can meet the engineering requirements.

基于多域物理信息神经网络的复合地层隧道掘进地表沉降预测

复合地层中盾构掘进诱发地表沉降的准确预测是隧道工程安全建设与施工决策的关键问题。基于隧道施工诱发地层变形机制构建隧道收敛变形与掘进位置的联系,并将其耦合至深度神经网络(deep neural network,简称DNN)框架,建立了预测盾构掘进诱发地层变形的物理信息神经网络(physics-informed neural network,简称PINN)模型。针对隧道上覆多个地层的地质特征,提出了多域物理信息神经网络(multi-physics-informed neural network,简称MPINN)模型,实现了在统一的框架内对不同地层的物理信息分区域表达。结果表明:MPINN模型高度还原了有限差分法的计算结果,可以准确预测复合地层中隧道开挖诱发的地表沉降;由于融入了物理机制,MPINN模型对隧道施工诱发地表沉降的问题具有普适性,可应用于不同地质和几何条件下隧道诱发地表沉降的预测;基于工程实测数据,提出的MPINN模型准确预测了监测断面的地表沉降曲线,可为复合地层下盾构掘进过程中地表沉降的预测预警提供参考。

高铁桥墩沉降的通用渐进分解长期预测网络模型

高铁桥墩不均匀沉降是导致轨道不平顺的潜在原因之一,准确预测桥墩沉降对于确保铁路建设和运营的可靠性和安全性具有重要意义。目前,常规时间序列领域的多数预测模型仅在预处理良好且没有缺失的数据集上进行测试,而在高铁桥墩沉降的真实场景中,沉降数据相较于其他领域存在观测频次少且不等时距,以及沉降规律复杂多变的问题,造成长期预测困难。为此,本文提出一种高铁桥墩沉降的通用渐进分解长期预测网络(GPDLPnet),摒弃传统的预处理思想,将预处理过程嵌入网络结构,在网络训练过程中实现渐进预处理。首先,GPDLPnet在每轮迭代中利用改进对角掩码自注意力模块分析沉降数据中的缺失模式。然后,通过改进完全自适应噪声集合经验模态分解模块将沉降数据分解并重构为高频、低频和趋势子分量,将子分量作为BiLSTM-RSA-Resnet预测模块的特征输入。最后,输出递归预测结果,从而实现高铁桥墩沉降的长期预测。结合实际工程数据,将数据划分为高频观测和低频观测两类典型的观测模式进行试验,在3~4个月的预测中GPDLPnet均表现出良好的预测性能,并在精度指标上优于其他7种模型。

基于多输入卷积神经网络隔震支座沉降识别

为了避免地基不均匀沉降导致隔震支座沉降以及对上部结构造成的隐性损伤,针对隔震支座沉降识别方法进行研究,提出一种基于多输入卷积神经网络(multi-input convolutional neural network,MI-CNN)的隔震支座振动信号识别模型。首先,采集隔震支座水平方向加速度和位移信号,采用归一化预处理和数据增强方法扩充样本;然后,将样本输入到所建立的网络模型中并进行训练;最后,利用完成训练的网络模型进行沉降识别。结果表明:相较于传统单输入卷积神经网络(Convolutional neural network,CNN)模型,MI-CNN模型易于训练,可最大程度地发挥CNN对沉降信号特征的提取能力,且具有更好的沉降位置识别准确率和更小的沉降程度识别误差,以及针对不均衡数据集更稳定的识别效果。研究结果可为隔震支座沉降识别提供新思路。

基于SSA-BP的深基坑地表变形预测研究

文中采用麻雀搜索算法优化BP神经网络,对深圳市某在建地铁车站深基坑周围地表沉降监测点进行变形预测。通过对基坑地表变形监测点DBC16-4的118期监测数据进行训练学习,并与粒子群算法优化BP神经网络、遗传算法优化BP神经网络和标准BP神经网络横向对比,验证了训练效果。结果表明,麻雀搜索算法对BP神经网络权重寻优速度较快,收敛精度更高,麻雀搜索算法优化BP神经网络模型预测平均相对误差仅为1.72%,拟合精度较其他算法更高,预测效果良好。

西南某机场跑道沉降预测模型

机场道面沉降,严重影响机场安全运行。准确预测跑道工后沉降,对机场的建设与运行极为重要。以西南某机场跑道沉降变形的观测数据为依据,分别用双曲线模型、对数模型、指数模型以及灰色预测模型,对跑道沉降进行预测和对比分析,解决了小样本数下曲线预测精度较低及灰色模型对非线性预测准确度差等问题,提高了预测的精度;同时通过BP神经网络对组合预测模型的残差进行修正,最大限度地提高模型预测的精度和效果,为地基沉降预测提供借鉴。

SBAS-InSAR技术融合CNN-LSTM模型的矿区开采沉陷监测与预测

针对传统矿区开采沉陷监测方法耗费人力财力和预测预警模型较少的问题,研究提出一种基于短基线集合成孔径雷达干涉测量(Small Baseline Subset-Interferometry Synthetic Aperture Radar,SBAS-InSAR)技术和卷积神经网络(Convolutional Neural Networks,CNN)与长短期记忆网络(Long Short-Term Memory,LSTM)相结合的矿区开采沉陷监测预测方法。首先,利用SBAS-InSAR技术对建新煤矿进行矿区开采沉陷监测,获取了该矿区的年平均沉降速率和累计沉降值。用GNSS监测数据与SBAS-InSAR结果进行对比验证,其拟合效果较好。其次,在此基础上利用CNN-LSTM模型预测后6期沉降数据,其结果与CNN和LSTM预测结果进行对比。研究显示,CNN-LSTM模型的平均绝对误差(S_(MAE))和均方根误差(S_(RMSE))比单一的CNN和LSTM分别至少降低了44.8%和40.6%,其决定系数均高于98%。最后,进一步预测前6期和中6期沉降数据,验证了CNN-LSTM预测模型在时间上的一致性。因此,SBAS-InSAR融合CNN-LSTM模型在类似矿山开采沉陷监测和预测中有较好的应用前景。

乡村聚落边界形态生成方法研究

聚落的边界由建筑部分的实边界与建筑之间空隙部分的虚边界连接而成。文章基于建筑外部空间理论与凸包原理,提出了一种求取聚落边界的新方法。首先利用Delaunay三角网求取聚落建筑之间的影响距离,继而绘制建筑节点网络图,以此求取聚落内两两之间具有直接可视空间关联的建筑之间的凸包,并叠加得出该聚落的公共空间,其外轮廓即为聚落的虚边界。综合实、虚两部分边界,得到完整的聚落边界图形。文章提出的新方法充分考虑了聚落内部建筑的疏密程度对聚落边界设定的影响,其结果更符合真实的空间感受,也更具科学性与合理性。

城市群流动人口居留意愿网络特征分析及空间模式划分

基于2017年CMDS数据,运用社会网络分析、地理空间分析方法,解析中国京津冀、长三角、珠三角、成渝、长江中游及哈长城市群的流动人口居留意愿网络空间结构差异,划分其空间模式。研究发现,城市群居留网络整体上呈不同等级的“圈层”结构,距离衰减效应显著。居留网络结构特征迥异,沿海城市群流动人口来源腹地广,京津冀和长三角集聚效应明显,流动人口居留意愿更高;珠三角城市群居留强度低,结构松散。内陆城市群吸引范围有限,居留意愿低,成渝城市群居留联系弱,内外部流动相对均衡,长江中游城市群呈现弱集聚与轻均衡特征,哈长城市群内部结构紧凑,外部联系松散。不同区位、不同规模、不同发展阶段的城市群其居留意愿网络空间格局差异明显,流动人口居留选择具有明晰的地域指向。沿海城市群应加大区域低中心性城市发展,创造更多的就业机会,以吸引人口流入;内陆城市群应优化产业结构布局,平衡人口流出与流入,避免区域人口过度流失。

基于遗传算法的BP神经网络在轻质路基沉降预测中的应用

为更好地掌握轻质路基施工过程中的沉降变形情况,选取宁芜保通线部分轻质路基沉降监测数据,在BP(back propagation)神经网络模型的基础上,采用遗传算法对其进行优化,并将优化后的模型应用于轻质路基沉降预测。结果表明:遗传算法优化的BP神经网络在全局搜索能力和收敛能力方面具有明显优势;在轻质路基沉降预测任务中,多数预测结果的相对误差集中在更低的范围内,监测点1和监测点2预测结果的模型评价指标MAE、RMSE、MAPE分别为0.017 mm、0.021 mm、0.679%和0.013 mm、0.016 mm、1.395%,预测结果拟合程度高,误差小,模型泛化能力强。因此,遗传算法优化的BP神经网络的沉降预测模型具有可靠的预测效果与预测精度,在实际工程中可行性较高,可作为轻质路基沉降预测和预警的一种辅助手段。

基于深度学习的盾构隧道施工地表沉降预测方法

针对现有盾构隧道施工引发地表沉降预测方法中存在的难以同时挖掘数据之间的非线性特征关系和双向时序信息的问题,通过融合卷积神经网络(CNN)、双向长短期记忆(BiLSTM)与自注意力机制(SA)提出一种基于深度学习的地表最大沉降预测方法(CNN-BiLSTM-SA)。该方法首先利用CNN提取网络输入数据之间的非线性特征关系,利用BiLSTM网络提取输入数据的双向时序信息,然后引入SA机制为CNN提取的特征分配相应的权重,有效捕获时间序列中的关键信息,最后通过全连接层输出最终地表沉降预测结果。以湖南万家丽路电力盾构隧道工程为依托构建地表沉降数据集,并选用ANN、RNN、LSTM、BiLSTM模型开展对比分析。研究结果表明:评估指标CNN-BiLSTM-SA的平均绝对误差(MAE)、均方根(RMSE)、决定系数(R2)、平均绝对百分误差(MAPE)均为最优,具有更好的地表沉降预测性能。

传统家族型聚落社会空间关联性研究——以东莞塘尾村为例

以广东省东莞市传统家族型聚落塘尾村为例,引入空间句法并以户为空间基本单元,定量描述空间的集聚效应、观察空间单元的结构性特征。弥补既有研究对空间的单元、界域、中心界定较模糊等问题。从社会网络和空间结构关联视角,探讨地域文化和宗族架构与空间的关联,族群内权力等级差异、职能调整和角色转变等对社会空间的影响以及外部社会事件和人物个性对空间的作用等。结合相关性分析,总结和推演以塘尾村为例的东莞地区传统家族型聚落社会空间转换模式及其影响机理,为更好地延续和振兴地方传统聚落提供新思路和新方法。

传统聚落社会网络演变特征及其空间关联机制研究

针对传统聚落出现的空心化、商业化现象,本文通过构建典型传统聚落社会网络模型揭示其近四十年的社会网络演变特征,探究社会运行逻辑与空间关联机制。研究显示:空心化传统聚落血缘关系由大家庭演变为隔代家庭以至空巢家庭结构,地缘关系由独户转向邻里到同乡结构,血缘地缘结构瓦解在一定程度上加速空间的衰败。商业化传统聚落血缘关系由大家庭演变为核心家庭以及混合家庭,业缘关系由同业关系转向合作及竞争结构,血缘结构瓦解与业缘结构复合过程中,传统聚落活动主体的社会经济活动发生改变,导致传统聚落的空间需求也相应改变,进而促成新的社会结构的形成与扩展。最后,尝试通过空心化、商业化传统聚落作用规律对比,为传统聚落物质空间与社会结构互动的保护找到切入口。

基于优化长短期记忆网络的矿坑遗产沉降预测

工业矿坑遗产以其独特风貌和价值逐步受到广泛关注。针对矿坑遗产易发的沉降地质灾害,积极采取预防措施是降低损失的有效途径。为解决工业矿坑遗产沉降灾害预测问题,提出一种融合蜣螂优化算法(DBO)的优化长短期记忆网络(LSTM)算法,用于构建预警模型。选取阜新市海州露天矿作为实验地点,利用小基线集合成孔径雷达干涉测量(SBAS-InSAR)技术采集55景矿区沉降数据。通过两种去噪方法对采集到的样本数据进行去噪处理,应用DBO算法优化LSTM,建立工业矿坑遗产沉降预测模型。LSTM模型的超参数使用DBO算法优化以实现高精度预测模型,并与其他算法优化LSTM后的模型指标进行对比。结果表明:DBO-LSTM模型在工业矿坑遗产沉降预测优势突出,预测模型的均方根误差、平均绝对误差和决定系数分别为0.045 mm,0.038 mm,0.956,均优于其他预测模型。DBO-LSTM模型在预测工业矿坑遗产沉降方面展现了高精度、快速收敛和强稳定性等特点,为工业矿坑遗产保护工作提供了有力支持。

基于TimeGAN增强的CNN-LSTM模型在盾构掘进地表沉降中的预测研究

为更准确地预测小数据量下盾构法施工造成的地表沉降,提出基于TimeGAN(time series generative adversarial networks,时间序列生成对抗网络)增强的CNN(convolutional neural networks,卷积神经网络)-LSTM(long short-term memory,长短期记忆网络)盾构掘进地表沉降预测模型,并依托上海北横通道新建工程Ⅱ标盾构施工项目验证该增强模型的性能。首先,选取300环的部分施工参数、地质参数、几何参数以及地表最大沉降,对比LSTM、CNN-LSTM与TimeGAN-CNN-LSTM的性能,证明CNN-LSTM对于盾构施工环境下多参数的预测效果明显优于LSTM,TimeGAN-CNN-LSTM增强模型优于CNN-LSTM;然后,通过更改训练集及测试集的大小,对不同数据集下TimeGAN-CNN-LSTM增强模型相较CNN-LSTM的预测效果进行研究。结果表明:TimeGAN-CNN-LSTM增强模型预测效果相较CNN-LSTM模型提升显著,且当训练集与测试集比值为4~8时,提升最为显著。