矿区岩溶地表塌陷神经网络预测模型研究

针对矿区岩溶地表塌陷存在的非线性动力学特征,为更准确预测岩溶地质矿区地表塌陷区域分布,在分析研究某矿区岩溶地表塌陷机理及其影响因素基础上,确定矿区地表塌陷的影响因素,构建矿区岩溶地表塌陷BP神经网络非线性动力学预测模型。采集并分析某岩溶矿区大量岩溶地表塌陷历史数据,应用Matlab神经网络工具箱,采用构建的矿区岩溶地表塌陷BP神经网络预测模型,对上述矿区岩溶地表塌陷区域分布情况进行非线性预测。研究结果表明,采用训练的神经网络预测模型可以实现对矿区岩溶地表塌陷危险性的合理预测。     

基于RFID技术的剧毒危化品物流服务平台研究

为了提高我国的剧毒危化品物流安全监管和信息化水平。结合剧毒品氰化钾的物流过程,提出一个基于RFID技术的剧毒危化品物流服务平台解决方案。研究的重点是适用于剧毒品氰化钾圆罐的主动式RFID电子标签研发,以及RFID中间件和基于B/S的应用系统开发。通过剧毒品物流服务平台的开发,表明该方案从数据采集、网络传输、智能应用3个层次解决剧毒品物流中监控和信息共享等问题,实现剧毒品仓储、运输过程安全监管和物流管理及服务的信息化。此外,该平台符合物联网建设的整体结构,其扩展能力强。     

Integrative trophic network assessments of a lentic ecosystem by key ecological approaches of water chemistry, trophic guilds, and ecosystem health assessments along with an ECOPATH model

The objective of this study was to describe the trophic structure and energy flow in a lentic ecosystem in South Korea. Physicochemical water conditions were evaluated along with the reservoir ecosystem health using a multimetric IBI model. Nutrient analyses of the reservoir showed a nutrient rich and hypereutrophic system. Guild analysis revealed that tolerant and omnivorous species dominated the ecosystem. Tolerant fish, as a proportion of the number of individuals, were associated (R² > 0.90, p < 0.01) with TN and TP, the key indicators of trophic state in lentic ecosystems. The mean Reservoir Ecosystem Health Assessment (REHA) score was 19.3 during the study, which was judged as in ‘fair to poor’ condition. A trophic analysis of the reservoir estimated by the ECOPATH model shows that most activity in terms of energy flow occurred in the lower part of the trophic web, where there was intensive use of primary producers as a food source. Consequently, of the 10 consumer groups, nine fell within trophic levels <2.8. Trophic levels (TL) estimated from the weighted average of prey trophic levels varied from 1.0 for phytoplankton, macrophytes, and detritus to 3.25 for the top predator, Pseudobagrus fulvidraco. Our integrated approach to trophic network analysis may provide a key tool for determining the effects of nutrient influx on energy flow pathways in lentic ecosystems.     

Quantifying the sustainability of water use systems: Calculating the balance between network efficiency and resilience

In ecological network theory, network efficiency and resilience are two essential but complementary attributes of the network structure, and a balance between these factors is critical for an ecosystem's long-term sustainability. Our paper introduces this method and related concepts into water use systems to provide a new angle for sustainability quantification. In this paper, we investigate the meanings of network efficiency and resilience in the context of sustainable development of water use systems, and define sustainable systems based on the optimal balance between network efficiency and resilience. With the consideration of complex artificial characteristics of water use, we propose an optimal water use network and quantify its flows. By ascendency calculation, the balanced network structure can be determined. We then use the four sub-basins of China's Haihe River as a case study to illustrate how the optimal network can be constructed and how the optimal balance for each scenario can be calculated. The results show that the optimal balance for the sub-basins has ascendency values ranging from 0.5970 to 0.7161. By analyzing the contribution of each water use activity to network's balance structure, the location of the optimal balance in water use systems can be better understood. This research represents the first attempt to explore the balance between a network structure's efficiency and resilience as a way to quantify the sustainability of water use systems, and builds a foundation for future studies on the assessment, regulation, and management of water resources.     

我国北方城市暴雨灾害演化过程及风险分析

提出一种基于灾害演化网络的风险分析方法,针对城市暴雨灾害的演化过程进行了风险分析。利用复杂网络理论构建北方城市暴雨灾害演化网络模型,将危机事件分为三个等级,并探讨了事件级别和出入度的关系;分析城市暴雨灾害链演化特点和暴雨危机事件后果蔓延规律,得到关键危机事件和演化链。结果表明,交通堵塞是我国北方城市暴雨灾害系统中关键危机事件;暴雨灾害演化系统存在短链、长直链和循环链三种结构,其中循环链中的危机事件互为因果,可自行激化,是灾害演化网络控制的关键结构。     

基于离散型Hopfield神经网络的台风灾情评估模型

台风灾情评估对防台减灾和救灾工作十分重要,为此,提出并建立了一种基于离散型Hopfield神经网络的评估模型。该模型选取5个评估因子,根据广东省台风灾害历史数据,按灾害等级划分原则制定了评估因子5个等级的划分方法,然后将待分级的台风进行Hopfield编码,建立离散型Hopfield神经网络模型,并用实例对模型进行了验证,分析模型的不足,将评估因子的权重因素引入到模型中,使模型更加合理。     

城市生命线系统的地震网络可靠性研究

震后生命线系统的网络可靠性对于灾后的应急救援和恢复至关重要,因此,评价城市生命线系统网络在一定地震强度下,网络系统保持连通的能力已成为研究生命线地震网络可靠性的重要内容。以经典的系统单元可靠性模型为基础,在众多的网络可靠性计算方法中根据方法实现的难易程度和计算效率,选择采用了不交化穷举集合法和有序二分决策图方法,比较研究了网络系统的可靠性问题,并与Monte Carlo方法进行了对照。此外,针对淮南市配水管网的实际案例,采用上述方法计算了相应的可靠度,计算结果表明,所提方法能够高效地应用于一般的工程实践,具有显著的实用意义。     

地下管网地震反应分析模型

建立了多点地震动激励作用下的地下管网地震反应拟静力分析模型.首先建立基于有限单元模型的地下管网系统刚度方程,接着给出考虑地震动空间相关性及场地条件等诸多因素的地震动场模拟思路及管网系统地震动施加策略,最后通过解析法验证了模型的正确性.采用本文方法对一“十”字形管网进行了地震反应分析,结果表明:①管线轴向应力沿着管线长度...     

改性水凝胶的制备及其对Pb2+、Cd2+吸附性能研究

采用60Co-γ射线低温辐射技术,以丙烯酸羟乙酯(HEA)和N-乙烯基甲酰胺(NVF)为共聚单体,合成了一种新型共聚物水凝胶:聚(丙烯酸羟乙酯/N-乙烯基甲酰胺)[p(HEA/NVF)],并采用三乙烯四胺(TETA)对其进行胺化改性,制得一改性水凝胶:聚(丙烯酸羟乙酯/N-乙烯基甲酰胺)-三乙烯四胺[p(H/V)-T].对制备的p(HEA/NVF)凝胶进行了SEM分析和FTIR分析,并对改性后的p(H/V)-T进行XPS分析及FTIR分析.研究了改性前后水凝胶对重金属Pb2+、Cd2+的吸附性能,改性后水凝胶对Pb2+、Cd2+的吸附动力学,以及重金属离子初始浓度、pH值等对改性凝胶吸附性能的影响,并对吸附重金属前后的水凝胶进行电子能谱分析.SEM分析表明,p(HEA/NVF)水凝胶是一种具有多孔结构的高分子材料;吸附实验结果表明,相比较p(HEA/NVF),p(H/V)-T对Pb2+的吸附量增加了700%,对Cd2+的吸附量增加了600%;p(H/V)-T对Pb2+、Cd2+的吸附量随着pH的升高而增加,吸附过程满足准二级动力学方程,并均能较好地符合Freundlich和Langmuir等温吸附模型(R2>0.95);红外分析证实,HEA和NVF已成功聚合为一新型共聚物水凝胶,且多胺基团已被成功地接于共聚物水凝胶表面;XPS分析得出,改性后水凝胶p(H/V)-T对重金属Pb2+、Cd2+的吸附机制为两者之间的螯合作用.     

基于深度卷积神经网络的场地污染非线性反演方法

提出了将Sobel边缘检测算子与深度卷积神经网络（CNN）算法结合的方法（E-ConvNet），用于污染场地的ERT反演过程.通过Sobel算子提取污染区域视电阻率数据的边缘特征作为CNN的先验信息，提高E-ConvNet的计算效率及识别精度.在5种理论模型（单异常体、双异常体及含双异常体的层状结构）和现场实例上测试了E-ConvNet算法的性能，并与最小二乘算法（LS）比较.测试结果表明：E-ConvNet能够准确识别污染处的面积、位置及阻值，其识别精度和计算效率均优于LS.E-ConvNet的单异常识别准确率为81.8%~84.9%，而LS则仅为9.6%~36.2%；多异常识别准确率为68.6%~84.4%，LS仅为2.8%~27.6%；E-ConvNet用时约为112~190ms，LS耗时为6000~7000ms.因此，在污染场地调查工作中，E-ConvNet能够准确高效地反演出污染区域的位置及范围，为开展后续评估/修复工作提供技术支持.