人工智能在短临降水预报中应用研究综述

短临降水预报是一项重要且具有挑战性的世界性难题.研究人员曾尝试使用各种技术预报降水,但是由于降水本身具有高度非线性、随机性和复杂性的特性,使得降水预测精确度并不高.近年来,随着人工智能技术的迅猛发展,其日渐渗透到人们生活的方方面面,气象领域也因此得益.人工神经网络能够对非线性系统进行建模,因此相比于传统方法,如数值天气预报法和光流法等,人工智能方法使得降水预报的准确率大大提高.本文介绍了传统降水预报的方法,着重总结概括了用于短临降水预报的各种最新人工智能方法,并对各研究方向进行归纳分析,为各类研究人员研究提供有益参考和借鉴.     

权重平衡有向网络下分布式约束优化的连续时间算法设计

本文研究权重平衡有向网络下分布式约束优化问题的求解,其中网络的全局目标函数是由每个智能体的局部目标函数的和构成,全局的约束是由每个智能体的局部约束的交构成.为了分布式求解该问题的最优解,首先引入智能体的局部共轭函数将其转换为Fenchel对偶问题.其次,从Fenchel对偶问题出发,提出一类基于奇异摄动系统的分布式连续时间算法.在局部目标函数和其梯度分别满足强凸和Lipschitz（李普希兹）连续的情况下,结合凸分析方法和Lyapunov（李雅普诺夫）稳定性理论,结果表明所提算法能够获得原问题和对偶问题的最优值.最后,数值仿真进一步验证了所提算法的有效性.     

强制通风温度传感器的计算流体动力学分析

本文利用计算流体动力学(Computational Fluid Dynamics, CFD)方法对基于076B型强制通风防辐射罩的温度传感器进行流固耦合传热分析,计算不同太阳辐射强度、下垫面反射率、强制通风速度、气象站海拔高度和下垫面长波辐射强度等影响因子条件下温度传感器的辐射误差。为实现影响因子连续变化时能提供准确修正数据的目标,采用神经网络算法仿真值计算结果进行拟合,获得辐射误差订正方程。结果表明,辐射误差订正方程的修正误差在0.005℃以内。订正后的气温观测数据有望用于气候变化研究和高精度天气预报。     

基于RBF神经网络的矿产资源潜力预测模型

提出一种基于RBF神经网络的矿产资源潜力制图模型。应用该模型生成矿产资源潜力分布图分三步完成：第一步,以找矿标志的空间分布图和已知矿点空间分布图为依据,提取训练样本;第二步,根据训练样本构建RBF矿产资源潜力制图模型;第三步,生成矿产资源潜力分布图。笔者以新疆北部阿尔泰多金属成矿带为研究区,比较了该模型与合成有矿可信度等模型的找矿靶区圈定结果。两种模型的靶区圈定结果基本相同,证明了RBF矿产资源潜力制图模型的有效性。     

海洋中溶解甲烷的原位检测技术研究进展

海水中溶解甲烷气体不但对全球变暖和海洋环境变化有着重要影响,而且也是发现渗漏型天然气水合物赋存区的依据之一,海水溶解甲烷原位监测的新技术和新方法是获取海水甲烷通量变化过程的主要手段。原位甲烷传感器具有原位、实时、便于多时空尺度定量观测等特点,在海洋环境变化和全球气候变化研究,以及海底资源开发利用中具有广泛的应用前景。介...     

青藏铁路路基下部土体水热过程变化的监测研究

修筑路基后土体水分变化是影响路基稳定性的重要因素,然而关于路基土体水分变化认识是极为缺乏的.采用了pF-meter传感器来监测土体温度和水分变化,获得了近3 a的连续监测数据.结果表明,pF-meter较好地反映了路基土体温度和水分变化特征以及土体水分迁移特征.但是,pF-meter传感器不能定量地给出土体水分变化,这...     

Use of artificial neural networks to identify the origin of green macroalgae

This study demonstrates application of artificial neural networks (ANNs) for identifying the origin of green macroalgae (Enteromorpha sp. and Cladophora sp.) according to their concentrations of Cd, Cu, Ni, Zn, Mn, Pb, Na, Ca, K and Mg. Earlier studies confirmed that algae can be used for biomonitoring surveys of metal contaminants in coastal areas of the Southern Baltic. The same data sets were classified with the use of different structures of radial basis function (RBF) and multilayer perceptron (MLP) networks. The selected networks were able to classify the samples according to their geographical origin, i.e. Southern Baltic, Gulf of Gdańsk and Vistula Lagoon. Additionally in the case of macroalgae from the Gulf of Gdańsk, the networks enabled the discrimination of samples according to areas of contrasting levels of pollution. Hence this study shows that artificial neural networks can be a valuable tool in biomonitoring studies.     

Carbon,nitrogen and phosphorus dynamics in nine sub-systems of the Sylt-Rømø Bight ecosystem,German Wadden Sea

Flow networks of nine sub-systems consisting of 59 components each of the Sylt-Rømø Bight, German Wadden Sea, were constructed depicting the standing stocks and flows of material and energy within and between the sub-systems. Carbon, nitrogen and phosphorous were used as currencies for each sub-system, thus resulting in 27 network models, which were analyzed by ecological network analytical protocols. Results show substantial variability in the dynamics of these elements within and between the nine sub-systems, which differ in habitat structure, species diversity and in the standing stocks of their constituent living and non-living components. The relationship between the biodiversity and selected information indices and ratios, derived from ecological network analysis, of individual sub-systems is variable and differ substantially between them. Ecosystem properties such as the structure and magnitude of the recycling of these elements, number of cycles, and total sub-system activity were calculated and discussed, highlighting the differences between and complexity of the flow of C, N and P in a coastal marine ecosystem. The average number of cycles increase from 179 for C, to 16,923 and 20,580 for N and P respectively, while the average amount of recycled material, as measured by the Finn Cycling Index (FCI), increase from 17% for C, to 52% for P and to 61% for N. The number of cycles and the FCI vary considerably between the sub-systems for the different elements. The largest number of cycles of all three elements was observed in the muddy sand flat sub-system, but the highest FCIs were computed for both C (32%) and N (85%) in the Arenicola Flats, and in sparse Zostera noltii sea grass beds for P (67%). Indices reflecting on the growth, organization and resilience of the sub-systems also showed considerable variability between and within the inter-tidal ecosystems in the Bight. Indices such as, for example, the relative ascendency ratios increase on average from C to N to P, whereas others, such as the Average Mutual Information and Flow Diversity indices, were found to be higher in the N models than in the C or P ones.     

基于GPRS无线传感网的大气CO2浓度采集系统研究与设计

针对实现大气中二氧化碳浓度的及时快速采集,设计了基于GPRS无线传感网的大气二氧化碳浓度采集系统,系统分为数据采集终端、GPRS无线传输网络和数据中心3部分.数据采集部分主要由单片机、高精度二氧化碳传感器、GPS模块、GPRS模块和液晶显示模块组成,负责采集二氧化碳浓度和采集点地理定位信息;无线传输网络部分由GPRS模块接入到GPRS网络,再登陆到Intemet网络与检测中心服务器建立TCP网络连接,负责将采集到的数据传输给数据中心;数据中心主要是接受GPRS网路传输过来的数据,并将数据进行处理、显示和录入数据库.采集系统能摆脱数据采集时地理位置和天气条件上的局限性,使检测工作变得更及时方便准确.     

气象预警信息综合发布平台的设计与实现

气象灾害监测预警及信息发布是气象工作的重要组成部分,预警信息综合发布平台以提高气象预警服务的效率及信息化水平为目的,应用数据库、软件编程及地理信息系统技术完成气象综合观测系统基础数据产品的获取及分析,确定气象灾害预警范围内的信息终端,高效编辑各类气象预警信息,基于通用分组无线服务、码分多址、Web Service等信息技术,实现了电子显示屏、气象预警机、决策手机等多种终端的预警信息在同一平台上的综合发布。