基于三角模糊数的机械伤害事故风险分析

为研究机械伤害事故风险因素的影响程度,更好地对机械系统进行安全管理,实现机械设备安全运行和生产,以某企业清理沥青修整机作业时发生的机械伤害事故为例对其进行风险评估,根据机械伤害事故的特点并结合《机械工厂安全性评价标准》构建了机械伤害风险评价指标体系,采用三角模糊数和网络分析法计算导致事故的各个因素和子因素的权重。结果表明,危险操作倾向、机器安全防护、身心状态是导致事故发生的直接因素,并据此提出了相应的风险预防措施,为机械伤害事故风险决策提供了有力依据。     

常州市燃气管网破坏的人工神经网络预测模型

作为一种高效清洁的能源,燃气已经成为城市能源中的重要一员,燃气管网破坏亦成为城市所面临的重大安全隐患。城市埋地燃气管网的破坏风险,往往受到多种影响因素的共同作用。通过分析常州市埋地燃气管网破坏的影响因素,确定了地面沉降、地裂缝、城市内涝、土壤腐蚀等4个风险评价因子。运用MATLAB中的人工神经网络工具,通过人工神经网络计算,优化了模型网络结构,建立了常州市埋地燃气管网破坏风险预测的人工神经网络模型。分析计算结果,并为常州市埋地燃气管网的安全防护提供了建议。     

基于YOLO-BP神经网络的古建筑修缮阶段火灾监测方法*

为提高古建筑修缮阶段火灾监测水平，有效预防火灾事故，提出1种基于YOLO-BP神经网络算法的古建筑修缮阶段火灾监测模型。针对修缮作业对古建筑产生的消防隐患，在施工场地设置监测点，并利用YOLO算法以火源、可燃物为目标进行检测与定位，计算火源与可燃物距离；以火源温度、含氧量、火源与可燃物距离3项指标作为BP神经网络输入层，以火灾概率为输出层，从而进行预测；基于古建筑修缮阶段焊接、切割动火作业阶段的样本数据进行仿真模拟。结果表明:利用模型监测古建筑修缮阶段火灾的准确率达93.9%，验证模型的可靠性；当火源温度在150~2 000 ℃范围内，含氧量不小于19.5%时，动火作业安全距离为10.1 m。     

计算机网络风险防范模式研究

网络风险管理的目的是为确保通过合理步骤,以防止所有对网络安全构成威胁的事件发生.计算机网络风险防范模式包括风险防范策略、风险防范措施和实施风险防范.网络风险防范策略主要有:①当存在系统漏洞时,实现保证技术来降低弱点被攻击的可能性;②当系统漏洞被恶意攻击时,运用层次化保护、结构化设计以及管理控制将风险最小化或防止这种情形的发生;③当攻击者的成本比攻击得到更多收益时,运用保护措施,通过提高攻击者成本来降低攻击者的攻击动机;④当损失巨大时,运用设计原则、结构化设计以及技术或非技术类保护措施来限制攻击的程度,从而降低可能的损失.风险防范措施包括风险假设、风险规避、风险限制、风险计划、研究和了解、风险转移.在实施风险防范的过程中包括对行动进行优先级排序、评价建议的安全控制类别、成本-收益分析、选择风险防范控制、分配责任、制定安全措施实现计划、实现被选择的安全控制及残余风险分析.最后以某市政府官方门户网络应用系统为例进行了实证.     

基于轴—辐理论的皖江城市带物流网络空间构建

大力发展现代物流业和皖江城市带承接产业转移示范区建设是两大重要国家级战略.首先对轴—辐理论的国内外研究现状进行了总结和归纳,分析了轴—辐网络的结构及其优劣势.其次,设计了物流中心性指数的评价指标体系,运用主成分分析法对皖江城市带的9个地级城市进行物流节点等级划分,根据分析结果的总得分和区位特征确定合肥、芜湖、安庆3个一级物流节点,形成轴三角.分析了各级物流节点间的干线及支线物流通道,构建了皖江城市带轴—辐物流网络空间.     

Using knowledge mapping to rethink the gap between science and action

Scholars have long stressed the need to bridge the gap between science and action and seek the most efficient use of knowledge for decision making. Many contributors have attempted to consider and understand the sociopolitical forces involved in knowledge generation and exchange. We argue, however, that a model is still needed to adequately conceptualize and frame the knowledge networks in which these processes are embedded. We devised a model for knowledge mapping as a prerequisite for knowledge management in the context of conservation. Using great ape conservation to frame our approach, we propose that knowledge mapping should be based on 2 key principles. First, each conservation network results from the conglomeration of subnetworks of expertise producing and using knowledge. Second, beyond the research-management gradient, other dimensions, such as the scale of operation, geographic location, and organizational characteristics, must also be considered. Assessing both knowledge production and trajectory across different dimensions of the network opens new space for investigating and reducing the gap between science and action.     

Different response of bacterial community to the changes of nutrients and pollutants in sediments from an urban river network

• Bacterial community varied spatially in sediments from the urban river network. • Key environmental factors shaping bacterial community were detected by RDA. • Bacterial co-occurrence networks changed at different levels of nutrient and metal. • Potential indicator species were selected to predict pollution risk in sediment. Microbial communities in sediment are an important indicator linking to environmental pollution in urban river systems. However, how the diversity and structure of bacterial communities in sediments from an urban river network respond to different environmental factors has not been well studied. The goal of this study was to understand the patterns of bacterial communities in sediments from a highly dense urbanized river network in the lower Yangtze River Delta by Illumina MiSeq sequencing. The correlations between bacterial communities, the environmental gradient and geographical distance were analyzed by redundancy analysis (RDA) and network methods. The diversity and richness of bacterial community in sediments increased from upstream to downstream consistently with the accumulation of nutrient in the urban river network. Bacterial community composition and structure showed obvious spatial changes, leading to two distinct groups, which were significantly related to the characteristics of nutrient and heavy metal in sediments. Humic substance, available nitrogen, available phosphorus, Zn, Cu, Hg and As were selected as the key environmental factors shaping the bacterial community in sediments based on RDA. The co-occurrence patterns of bacterial networks showed that positive interaction between bacterial communities increased but the connectivity among bacterial genera and stability of sediment ecosystem reduced under a higher content of nutrient and heavy metal in average. The sensitive and ubiquitous taxa with an overproportional response to key environmental factors were detected as indicator species, which provided a novel method for the prediction of the pollution risk of sediment in an urban river network.     

基于多源辅助数据和神经网络模型的稻田土壤砷空间分布预测

合适的方法和多源的辅助数据对于准确预测土壤重金属的空间分布具有重要意义.该研究提出一种径向基函数神经网络结合普通克里格法的模型(RBFNN_OK),由主成分分析(PCA)提取的地形因子、遥感数据和邻近信息等多源辅助数据作为自变量,预测江西省都昌县稻田土壤砷空间分布.为验证RBFNN_OK的可行性:首先在全县范围内采集144个稻田表层(0～20 cm)土壤样品,运用ArcGIS地统计模块随机抽取115个(80%)采样点作为测试集,29个(20%)采样点作为验证集.其次多源辅助数据包括地形因子、遥感数据和邻近信息等14个定量因子作为预测变量,将预测变量进行主成分分析,得到前10个主成分的累积贡献率达到97.62%.再次一个特定的RBFNN_OK被用来预测土壤砷空间分布.最后将RBFNN_OK模型的预测结果与径向基神经网络模型(RBFNN)、回归克里格模型(RK)和多元逐步线性回归模型(MSLR)进行比较.结果表明,RBFNN_OK的测量值标准偏差与均方根误差的比值(RPD)较其它3种方法分别提高了14.92%、35.71%和44.67%.此外,RBFNN_OK还提供了更加真实且有关土壤砷空间分布的细节信息.RBFNN_OK取得最优效果可能归因于引入多源辅助数据,考虑多源辅助数据和土壤砷之间的多重共线性和非线性关系.该方法可为稻田土壤砷调查与环境保护提供更为精准的信息.     

多源数据融合的市县国土空间规划人口城镇化模式——以扬州市为例

围绕“人、地、财”等新型城镇化三大核心问题,衍生出人口流动的社会、空间、经济属性,构建多源数据融合的人口城镇化概念模型。基于迁移流、位置识别和企业关联网络等方法,展开实证研究。结果表明：（1）区域层面：受落户门槛、家庭权益、社会保障等限制,人口流动比资本流动更为谨慎,也更遵从行政区经济和城镇等级体系。（2）市域层面：城镇化形式主要表现为城市和乡村的直接联系,重点镇辐射和吸纳作用不强。（3）新经济使后发地区弯道超车成为可能。研究发现此概念模型适用性较好,能够合理识别经济社会联系的主体格局,为城镇开发边界划定和国土空间格局优化起到一定的支撑作用,对市县国土空间规划的区域协调研究具有较强的参考价值。