基于SSA-MGF的偏最小二乘回归神经网络的预报模型

本文采用奇异谱分析（Singular Spectrum Analysis,SSA）方法对原始降水序列重构,并用均生函数（Mean Generating Function, MGF）方法对重构系列构造延拓矩阵,以此作为自变量,原始降水序列作为因变量, 再利用偏最小二乘法提取对因变量影响强的成分作为神经网络的输入因子,原始序列作为输出因子,建立神经网络预测模型.通过对广西全区6月份降水量进行实际建模并与其它方法进行对比预测试验,结果表明,基于SSA-MGF的偏最小二乘回归神经网络预测模型较好,是一种具有较高应用价值的预测方法.     

强化环境质量监测 推进监测事业科学发展

从强化环境质量监测、评价与研究的重要性角度,简述了构建完善的由纵向网、立体网、横向网等3大网络体系构成的环境质量监测预警体系的必要性,提出该体系的构建需尽快完善法律、人才、经费、技术等4大长效保障机制。     

A fuzzy analytic network process for multi-criteria evaluation of contaminated site remedial countermeasures

The Analytic Network Process (ANP) has been proposed to incorporate interdependence and feedback effect in the prioritization of remedial countermeasures using a hierarchical network decision model, but this approach seems to be incapable of capturing the vagueness and fuzziness during value judgment elicitation. The aim of this paper is to present an evaluation method using a fuzzy ANP (FANP) approach to address this shortcoming. Triangular fuzzy numbers (TFN) and their degree of fuzziness are used in the semantic scale as human judgment expressed in natural language is most often vague and fuzzy. The method employs the alpha-cuts, interval arithmetic and optimism index to transform the fuzzy comparative judgment matrix into set of crisp matrices, and then calculates the desired priorities using the eigenvector method. A numerical example, which was drawn from a real-life case study of an uncontrolled landfill in Japan, is presented to demonstrate the process. Results from the sensitivity analysis describe how the fuzziness in judgment could affect the solution robustness of the prioritization method. The proposed FANP approach therefore could effectively deal with the uncertain judgment inherent in the decision making process and derive the meaningful priorities explicitly from a complex decision structure in the evaluation of contaminated site remedial countermeasures.     

Disaster management and mitigation: the telecommunications infrastructure

Among the most typical consequences of disasters is the near or complete collapse of terrestrial telecommunications infrastructures (especially the distribution network–the 'last mile') and their concomitant unavailability to the rescuers and the higher echelons of mitigation teams. Even when such damage does not take place, the communications overload/congestion resulting from significantly elevated traffic generated by affected residents can be highly disturbing. The paper proposes innovative remedies to the telecommunications difficulties in disaster struck regions. The offered solutions are network-centric operations-cap able, and can be employed in management of disasters of any magnitude (local to national or international). Their implementation provide ground rescue teams (such as law enforcement, firemen, healthcare personnel, civilian authorities) with tactical connectivity among themselves, and, through the Next Generation Network backbone, ensure the essential bidirectional free flow of information and distribution of Actionable Knowledge among ground units, command/control centres, and civilian and military agencies participating in the rescue effort.     

Least-inference methods for constructing networks of trophic flows

The construction of material and energy budgets within ecosystems has long been accomplished via manual calculation. Recently, optimization techniques have been adapted to automate the procedure, but these methods require assumptions that may not square with biological reality. Two algorithms are developed to construct ecosystem budgets under minimal inference. Although the methods do not recapitulate the model used to generate the input data, analysis reveals that the results do not differ statistically from networks that were constructed manually.     

AN EVALUATION OF TWO HYDROLOGIC MODELS FOR CLIMATE CHANGE SCENARIOS1

ABSTRACT: Understanding the effects of climate change on water resources requires coupling atmospheric and hydrologic models. With the wide array of hydrologic models, from simple empirical to complex physically based, it is not clear which is preferable to simulate hydrologic variations over long time scales. To address this issue, a black-box artificial neural network (ANN) model was compared to a distributed parameter conceptual Geographic Information System based Hydrologic Modeling System (GIS-HMS). Both models computed daily direct surface runoff in four sub-basins of the West Branch of the Susquehanna River Basin, Pennsylvania and were evaluated with five objective functions. Overall, results were comparable between models. However, the ANN was favored in the larger sub-basins, while GIS-HMS was more accurate in the smaller catchments. Both models were impaired by the poor spatial and temporal resolution of precipitation data and the simplified representation of antecedent soil-moisture conditions. In the context of climate change, where simulations are limited by computing power, results suggest that both models are appropriate. When detailed simulations are essential, GIS-HMS is a preferable model to use. On the other hand, the ANN model is more suitable when multiple scenarios require immediate analysis and the distributed qualities of runoff are not required.     

基于Abaqus软件的并行计算集群平台构建与优化方法

根据有限元显式算法和隐式算法的特点,研究了岩土工程动力分析并行计算集群系统的硬件要求、集群系统的构建方法,构建了基于EM64T硬件构架、双路Intel Xeon处理器、Linux操作系统和64位Abaqus软件的32CPU并行计算集群平台,测试了存储子系统对集群性能的影响,比较了两种千兆以太网络作为集群子网络的性能优化方法。以地下结构的地震反应分析为例,测试了优化前后该集群系统的计算速度,发现两种以太网络性能优化方法都可以有效提高集群计算速度。列举了该集群系统在深水桥梁基础流固耦合动力分析、地下结构地震反应分析和快速轨道交通环境振动分析中的应用,显示了该集群在显式算法、隐式算法及小规模、大规模数值计算问题中的并行计算效率,证明所构建的Abaqus数值模拟并行计算集群平台能够满足计算规模、计算精度和时效性的要求。     

旱涝气候监测、预警分析业务系统及服务

旱涝监测、预警分析业务系统系应用ＦＯＲＴＲＡＮ７７,ＢｏｒｌａｎｄＣ和Ｆｏｘｐｒｏ等语言混合编写而成,系统软件结构合理,界面友好,操作灵活方便。该系统建立了基本气候资料、旱涝指标和旱涝灾情等数据库,实现了实时气象资料接收,旱涝信息的统计、检索、列表显示和图形分析显示等功能,能实时、有效地监测、预警旱涝气候灾害的发生、发展,在为有关部门提供抗旱防洪决策科学依据中发挥了重要的作用。     

火山玻璃的拉曼光谱研究

本文分析了浙、闽、赣地区产于玄武岩、安山岩、流纹岩、黑曜岩及珍珠岩中玻璃相的拉曼光谱特征,认为酸性火山岩中的玻璃相分子团结构基本上由类似于结晶学上的、但在三维方向完全无序的链状、层状、架状及纯架状网络体组成,而中基性火山岩中的玻璃相结构仅由前三者组成。玻璃相分子团结构与其断健程度(DBB)间受一维延长的链状网络体(Ch)的含量正向控制及三维扩展的纯架状网络体(Ne~*)的含量负向控制,一元回归线性方程、相关系数R、拟合度F,分别是: 经研究,其中的黑曜岩可能是岩浆喷发间歇后的另一次爆溢产物。     

The “data-rich but information-poor” syndrome in water quality monitoring

Water quality monitoring conducted routinely over time at fixed sites has been a part of most water quality management efforts for many years. It has been assumed that such monitoring plays a major role in management. However, the lack of routine data analysis, and reporting of information derived from such analysis, points up the fact that the exact nature of the role of routine, fixed-station monitoring is poorly defined.There is a need to very clearly define this role in the design of such systems if routine monitoring is to efficiently and effectively meet the information expectations placed on it. Design of routine monitoring systems will therefore have to consider not only the where, what, and when of sampling, but also why. A framework for including the why of monitoring in the design process is proposed and experience with using the framework in New Zealand is discussed.