苏皖交界河网区浮游植物群落结构及其与环境因子关系的研究

苏皖交界处的石臼湖是长江下游唯一的通江淡水湖,通过青弋江和姑溪河直接与长江相连,形成了复杂的河网水系。为探讨这种复杂的河网水系中浮游植物种群结构特征,2012年平水期和枯水期分别对石臼湖及其周边入湖支流进行了浮游植物调查,研究河网水系中河网与湖区浮游植物的种类组成、季节变化及与环境因子的关系。结果显示:共采集到浮游植物105种,平水期与枯水期种类数差别不大,两次调查均出现的种类约有70%相同;浮游植物丰度最高可达1×108cell/L,达到了水华暴发的标准。优势种类主要绿藻门和蓝藻门为主,但是从生物量来看硅藻门占优。排序分析显示河网区域浮游植物季节演替明显,且河网与湖区浮游植物群落结构差异明显,进一步分析则表明影响浮游植物主要的环境因子是氮、磷等营养元素。     

Evaluating efforts to build sustainable WEEE reverse logistics network design: comparison of regulatory and non-regulatory approaches

Many developing countries such as Turkey are still making an effort on building an infrastructure for waste of electrical and electronic equipment (WEEE) reverse logistic network design (RLND) processes. It is obvious that policies/laws/regulations related to WEEE management provide a sustainable framework for implementation in the RLND. The question is here: Does the implementation of WEEE directives make sense in terms of reducing the total cost of the network in the long term? This study aims to compare regulatory and non-regulatory situations of WEEE RLND in developing countries by formulating two models named as ‘regulatory’ and ‘non-regulatory’. Model 1 is considered as sustainable with economic, environmental and social goals, and the quotas imposed by the environmental directive are taken into consideration as the data of product return amount. In Model 2, only economic goal is considered, and product return amount is forecasted using Artificial Neural Network (ANN). A case study is conducted in a recycling company in order to evaluate performance of the proposed models. This study contributes to the relevant literature by (1) comparing the regulatory and non-regulatory situations RL models explicitly and (2) proposing ANN model to forecast EEE product return or WEEE quantity for non-regulatory situation.     

基于神经网络模型的房屋震害易损性估计方法

结构易损性分析是震害预测工作中的一个重要环节.本文提出了一种基于人工神经网络模型的房屋震害易损性估计方法,文中以多层砖房结构的建筑为例,给出了易损性估计神经网络模型的输入向量设计方案和具有输出约束控制的BP神经网络模型.     

江苏省数字地震台网在考核运行期间的地震编目

江苏省数字地震台网从 1999年 6月至 9月记录到的自动触发和人工截取的 4 13条事件中 ,编报出 94个地震事件 ,其中 ,可用于数字网和模拟网进行地震三要素对比的地震事件有 35个。最终产出一份完整的地震目录和台网观测报告。进而从地震编目角度对数字地震台网资料和模拟台网资料确定地震三要素的可衔接性进行了尽可能的对比论证。     

Evaluating outcomes of conservation with multidimensional indicators of well-being

Many conservation interventions are hypothesized to be beneficial for both the environment and people's well-being, but this has rarely been tested rigorously. We examined the effects of adoption or nonadoption of a conservation intervention on 3 dimensions of people's well-being (material, relational, and subjective) over time. We focused on a fisheries bycatch management initiative intended to reduce environmental externalities associated with resource extraction. We collected panel data from fishers (n = 250) in villages with (adopters and nonadopters) and without (control) the conservation intervention 3 times over 2 years. We found no evidence that adoption reduced any of the 3 dimensions of well-being in the local populations affected by the intervention. There were modest improvements in material (t = –1.58) and subjective livelihood well-being (p = 0.04) for adopters relative to nonadopters over time. The variations in well-being experiences (in terms of magnitude of change) among adopters, nonadopters, and controls across the different domains over time affirmed the dynamic and social nature of well-being.     

斜坡稳定性评估神经网络专家系统N^2ES^3E

在于工程实列,采用神经网络理论,建立了斜坡稳定性评估分支结构专家系统,并针对斜坡评价的不同阶段,分别建立了以地质信息为３指标和斜坡的几何与力学参数为评价指标的两种评估系统。文中阐述了该系统的开发设想,介绍了系统的主要组成部分,同时给出了应用实例,应用结果表明该系统是有效的。     

受HDPE膜影响下的垃圾填埋场渗滤液水位探测方法研究

渗滤液水位会影响填埋场堆体稳定并有渗漏污染风险,当渗滤液赋存于HDPE防渗膜上方,堆体和渗滤液堆积电阻率特征的极端分异特性以及边界效应等因素使得最小二乘（LS）等传统物探反演方法无法精确反演实际电阻率分布,从而无法根据电阻率差异特征定位HDPE膜上方渗滤液水位的高度。为准确刻画堆体内部特别是渗滤液-HDPE膜局部电阻率精细分布,对传统高密度电法(ERT)装置进行改进,提出了一种川形探测装置(C-ERT),并采用BP神经网络的电阻率反演模型算法。通过COMSOL理论模型和江西某生活垃圾填埋场采集的现场数据对该方法进行验证,并与LS法比较。结果表明：基于川形装置的BP神经网络能有效识别出HDPE膜上方的渗滤液区域,识别准确率约为83.2%,而LS法并不能识别出渗滤液区域。

     

米诺环素对活性污泥初期增殖阶段的影响

采用污泥连续培养的方式,分别考察了米诺环素在质量浓度为0、0.001、0.01、0.1、1和10 mg·L⁻¹水平下对活性污泥初期增殖阶段的影响。结果表明,各组污泥增殖均能在20 d内进入平台期,且组间的稳态污泥量差异不显著 (P>0.05),表明活性污泥宏观生物量的增长与米诺环素浓度无关,但10 mg·L⁻¹米诺环素的暴露可抑制污泥絮体的形成。细菌群落多样性的分析结果表明,1 mg·L⁻¹的米诺环素即可导致活性污泥初期增殖阶段细菌多样性的降低,10 mg·L⁻¹的米诺环素会选择性富集Flavobacterium菌属,使其丰度在第20天的活性污泥细菌群落中达到83.14%,可进一步降低形成的活性污泥细菌多样性。网络分析结果表明,Proteobacteria细菌是活性污泥初期增殖阶段的关键物种,低质量浓度米诺环素(0.001 mg·L⁻¹和0.01 mg·L⁻¹)可能促进了活性污泥细菌间的互作,而高质量浓度(1 mg·L⁻¹和10 mg·L⁻¹)暴露会显著降低互作网络的复杂度和稳定性。本研究结果可为含抗生素废水的高效生化处理提供参考。     

基于人工神经网络的居民生活垃圾可燃成分热值预测

研究采用BP、RBF和自适应神经模糊推理系统(ANFIS)对生活垃圾可燃成分的热值进行预测。结果表明,BP神经网络模型的预测准确率为93.36%,RBF模型为96.87%,ANFIS模型为91.06%,3种模型均可用于可燃成分热值预测,但RBF模型的预测准确率相对较高,更适用于可燃垃圾的热值预测。     

AN EVALUATION OF PRECIPITATION GAGE DENSITY IN A MOUNTAINOUS TERRAIN1

ABSTRACT: Inter-station analysis was employed to evaluate the adequacy of the precipitation network in topographically complex West Virginia. A 25-year period was determined as the minimum lingth of record needed for relatively stable and fairly accurate estimates of long-term (50-year) precipitation and in frequency analysis. Data from the 83 National Weather Service stations with 25-year records were adjusted for consistency and evaluated separately by zones east (31 stations) and west (52 stations) of the Appalachian divide. Correlation coefficients (r) and average standard errors of estimate were computed for all station pairs within 50 miles distance and 1000 feet elevation difference of each other. The third polynomial equation of inter-station distance eliminated using elevation and land slope as the criteria in network design in this mountainous terrain. A network with (r) = 0.9 estimates annual precipitation with accuracy as great as 5 percent, but requires about 250 additional gages (i.e., about 200 percent of the present density).