基于广义极值分布的流化床压力波动风险预警

为预警气泡运动所引起的流化床粉煤气化压力波动风险,提出预测压力波动极值以及压力波动重现水平的方法;首先采用自相关分析法将压力波动母样本数据合理分段,再用区间极值法统计子样本的压力波动极值数据,以广义极值(GEV)分布方法建立GEV分布模型和Gumbel分布模型,并经过模型诊断选择最优模型;然后通过子样本与母样本的参数关...     

基于复合应力加速寿命试验的高速列车电子产品可靠度计算方法研究

目的 解决目前加速寿命试验一般不用于估计产品可靠度,而采用可靠性预计和应力-强度模型计算可靠度时可能存在因参数误差而影响其估计精度的问题.方法 提出一种基于加速寿命试验的高速列车电子产品使用可靠度计算方法,采用温度、湿度、振动综合应力和广义Egring模型对高速列车电子产品开展加速寿命试验.结果 通过试验数据可评估出产...     

主元分析-神经网络岩爆等级预测模型

为准确可靠地预测岩爆灾害,构建结合主元分析法(PCA)的径向基神经网络(RBFNN)、概率神经网络(PNN)和广义回归神经网络(GRNN)岩爆预测模型.选取6个常用的参数构成岩爆预测指标体系,采用PCA消除各指标间的相关性并降维,得出3个线性无关的主元即岩爆综合预测指标Y1、Y2和Y3,构成RBFNN、PNN、GRNN...     

超大城市昼夜交通事故严重程度致因对比分析——以深圳市为例

为研究夜间交通事故严重程度致因,基于深圳市3年3 244起交通事故数据,获取昼夜交通事故分布的时空特征;进一步选取交通事故集聚的南山区、福田区、罗湖区的1 798起交通事故,以交通事故严重程度为因变量,以事故原因、日期、事故形态等10个因素为候选自变量,构建广义有序Logit回归模型,对比分析昼夜不同严重程度交通事故的影响因素。结果表明:路口路段类型、疲劳驾驶、事故日期在夜间模型参数估计值分别为0.493,-0.363,-0.309,而在日间模型表现为不显著,道路路面材料在日间模型参数估计值为-0.232,而在夜间表现为不显著;事故原因、道路横断面渠化方式等因素在日间和夜间所引起交通事故的严重等级均存在较大差异。     

Ensemble Streamflow Forecast: A GLUE‐Based Neural Network Approach1

Abstract: While training a Neural Network to model a rainfall‐runoff process, generally two aspects are considered: its capability to be able to describe the complex nature of the processes being modeled and the ability to generalize so that novel samples could be mapped correctly. The general conclusion is that, the smallest size network capable of representing the sample distribution is the best choice, as far as generalization is concerned. Oftentimes input variables are selected a priori in what is called an explanatory data analysis stage and are not part of the actual network training and testing procedures. When they are, the final model will have only a “fixed” type of inputs, lag‐space, and/or network structure. If one of these constituents was to change, one would obtain another equally “optimal” Neural Network. Following Beven and others' generalized likelihood uncertainty estimate approach, a methodology is introduced here that accounts for uncertainties in network structures, types of inputs, and their lag‐space relationships by looking at a population of Neural Networks rather than target in getting a single “optimal” network. It is shown that there is a wide array of networks that provide “similar” results, as seen by a likelihood measure, for different types of inputs, lag‐space, and network size combinations. These equally optimal networks expose the range of uncertainty in streamflow predictions and their expected value results in a better performance than any of the single network predictions.     

山区沿河公路地质风险形成机制

根据山区沿河公路灾害的特点,将以岩土介质为主的传统地质灾害类型和水沙介质灾害类型整合,提出了“广义地质灾害体”的概念,并以介质组成为标准进行了广义地质灾害体分类.在孕灾环境评价和致灾因子评价的基础上,提出了广义地质灾害体的孕灾环境与致灾因子异相耦合发育机理.基于公路承灾体健康的理念,认为公路承灾体易损性主要受控于公路结构本身的健康性态,采取技术可行、经济合理的技术措施使结构健康复原是工程性减灾措施的目的.将公路承灾体类型分为结构性承灾体和功能性承灾体两类.提出了公路地质风险的耦合对抗形成机制,即公路地质风险是广义地质灾害体危险性与公路承灾体易损性之间时空耦合对抗的结果.认为地质风险评估是关于多因素非线性灾害风险系统的预测评价问题.根据地质风险形成的耦合机制,提出了采用解耦措施来逆向控制公路地质风险形成演化过程的减灾思路.针对山区沿河公路,给出了从孕灾环境、致灾因子、广义地质灾害体危险性、承灾体易损性评价到地质风险评估的思路与一般函数表达式.     

基于GAMs模型分析成都市气象因子交互作用对O3浓度变化的影响

为探究成都市大气环境中气象因子交互作用对臭氧(8h浓度平均最大值,统一用O₃表示)浓度变化的影响特征,利用成都市2014～2019年逐日大气污染物资料以及同期的气象资料,采用广义相加模型(generalized additive models,GAMs)分析气象因子对O₃浓度变化的影响效应.结果表明,单影响因素的GAMs模型中,O₃浓度与最高气温、日照时数、相对湿度、风速、降水量、最大混合层厚度(maximum mixed depth,MMD)和通风系数(ventilation coefficient,VC)间均呈非线性关系,无论全年还是夏季,最高气温、日照时数、MMD和相对湿度对O₃浓度影响均较大,值得注意的是,夏季相对湿度和降水量对O₃浓度变化的影响较全年更加显著.在构建O₃浓度变化的多气象因子GAMs模型中,除平均风速以外的其他气象因子共同作用对O₃浓度变化有显著影响,就全年而言,构建的GAMs模型判定系数(R²     

骆马湖浮游植物演替规律及驱动因子

江苏骆马湖作为南水北调工程的重要调蓄湖泊,水生态系统结构变化不容忽视.为研究骆马湖浮游植物群落结构演替规律及其与环境因子的关系,于2014～2018年进行了逐月监测.研究期间骆马湖总氮、高锰酸盐指数、电导率等参数呈逐年升高趋势,氟离子浓度呈逐年下降趋势.共鉴定有浮游植物7门71属,月均生物量变化范围0.16～5.51mg·L^-1.硅藻和绿藻为绝对优势门类,其次为甲藻及隐藻;主要优势属为针杆藻(Synedra sp.)、隐藻(Chroomonas spp.)、直链硅藻(Aulacoseira spp.)、锥囊藻(Dinobryon sp.)、栅藻(Scenedesmus spp.)、脆杆藻(Fragilaria spp.)、转板藻(Mougeotia sp.)、纤维藻(Ankistrodesmus sp.)和裸藻(Euglena spp.). 2014～2018年骆马湖浮游植物群落结构年际差异较大,其变化主要体现在浮游植物生物量的再分配,硅藻和绿藻继续保持优势外甲藻和蓝藻的优势度增加.非度量多维尺度分析显示,骆马湖浮游植物群落变化与总氮、氟离子、水温、总磷、溶解氧、...     

Cost‐Effectiveness of Using Small Vertebrates as Indicators of Disturbance

In species‐rich tropical forests, effective biodiversity management demands measures of progress, yet budgetary limitations typically constrain capacity of decision makers to assess response of biological communities to habitat change. One approach is to identify ecological‐disturbance indicator species (EDIS) whose monitoring is also monetarily cost‐effective. These species can be identified by determining individual species’ responses to disturbance across a gradient; however, such responses may be confounded by factors other than disturbance. For example, in mountain environments the effects of anthropogenic habitat alteration are commonly confounded by elevation. EDIS have been identified with the indicator value (IndVal) metric, but there are weaknesses in the application of this approach in complex montane systems. We surveyed birds, small mammals, bats, and leaf‐litter lizards in differentially disturbed cloud forest of the Ecuadorian Andes. We then incorporated elevation in generalized linear (mixed) models (GL(M)M) to screen for EDIS in the data set. Finally, we used rarefaction of species accumulation data to compare relative monetary costs of identifying and monitoring EDIS at equal sampling effort, based on species richness. Our GL(M)M generated greater numbers of EDIS but fewer characteristic species relative to IndVal. In absolute terms birds were the most cost‐effective of the 4 taxa surveyed. We found one low‐cost bird EDIS. In terms of the number of indicators generated as a proportion of species richness, EDIS of small mammals were the most cost‐effective. Our approach has the potential to be a useful tool for facilitating more sustainable management of Andean forest systems. Rentabilidad del Uso de Pequeños Vertebrados como Indicadores de Perturbaciones     

Contribution of Urban Expansion and a Changing Climate to Decline of a Butterfly Fauna

Butterfly populations are naturally patchy and undergo extinctions and recolonizations. Analyses based on more than 2 decades of data on California's Central Valley butterfly fauna show a net loss in species richness through time. We analyzed 22 years of phenological and faunistic data for butterflies to investigate patterns of species richness over time. We then used 18–22 years of data on changes in regional land use and 37 years of seasonal climate data to develop an explanatory model. The model related the effects of changes in land‐use patterns, from working landscapes (farm and ranchland) to urban and suburban landscapes, and of a changing climate on butterfly species richness. Additionally, we investigated local trends in land use and climate. A decline in the area of farmland and ranchland, an increase in minimum temperatures during the summer and maximum temperatures in the fall negatively affected net species richness, whereas increased minimum temperatures in the spring and greater precipitation in the previous summer positively affected species richness. According to the model, there was a threshold between 30% and 40% working‐landscape area below which further loss of working‐landscape area had a proportionally greater effect on butterfly richness. Some of the isolated effects of a warming climate acted in opposition to affect butterfly richness. Three of the 4 climate variables that most affected richness showed systematic trends (spring and summer mean minimum and fall mean maximum temperatures). Higher spring minimum temperatures were associated with greater species richness, whereas higher summer temperatures in the previous year and lower rainfall were linked to lower richness. Patterns of land use contributed to declines in species richness (although the pattern was not linear), but the net effect of a changing climate on butterfly richness was more difficult to discern. Contribución de la Expansión Urbana y un Clima Cambiante a la Declinación de la Fauna de Mariposas