基于粒子群算法的模糊优选神经网络储层识别模型

针对致密储层中气水干层识别准确率较低这一难题,提出基于粒子群算法(particle swarm optimization,PSO)的模糊优选神经网络识别模型.其基本思路:首先对所有的变量进行分析,提取反映储层本质的主要属性,消除重叠信息的输入,然后将粒子群算法与模糊优选神经网络相结合,构建基于粒子群算法的模糊优选神经网络识别模型.以陕甘宁盆地中部气田马五1储层气水干层识别问题为例,选用十九口井分层测试的92个已知样本(其中八十个用于模型构建,十二个用于精度检验),对物性、测井和储渗特征等方面的十个特征参数进行分析,从中提炼出电阻率、自然伽玛、产能系数、储渗因子、介质类型因子等五个主成份控制特征参数,消除重叠信息的影响.并以此作为神经网络输入层的输入,以样本储层的产能赋值作为输出,构建基于粒子群算法的模糊优选神经网络模型.模型的识别正确率达到100%,标准误差比传统模糊神经网络降低了60%.这表明该模型具有更高的识别精度,为致密储层的准确识别探索了又一种新方法,对同类地区的研究在一定程度上具有指导作用.     

贵州省德江县耕地硒含量特征和富硒农作物研究

为获得机器人听觉行为隐私感知方法,解决语音监听设备存在的隐私泄露风险,本文提出了基于卷积神经网络的服务机器人听觉隐私信息分类算法(APICA)。首先,设计了基于卷积神经网络的服务机器人听觉隐私信息分类算法及其卷积神经网络模型;其次,给出了机器人的听觉隐私信息监听系统工作流程;最后,为评估该听觉隐私信息分类算法性能,构建了训练和测试数据集,并在服务机器人平台上部署和实现了该算法。测试结果表明：系统识别隐私信息的平均精确率P、召回率R和F1值分别为96.35%、93.20%和94.53%,具有良好的识别和分类效果。     

径流预报的模糊神经网络方法

径流预报是水资源系统优化调度的重要因素。探讨了神经网络及模糊模式识别的部分理论,在通常的神经网络预报模型及模糊模式识别预报模型的基础上提出了一种模糊模式识别神经网络预报模型及其相应算法,加强了系统的知识表达能力,使预报结果更为可信。通过实例计算,验证了模型的可行性以及所给算法对模型训练的有效性,从而为径流预报提供了一种新方法。     

基于粒子群优化的细胞神经网络油气重力异常信息提取

细胞神经网络方法是一种有效的重力异常提取方法,提取出的重力异常信息具有较强的横向分辨能力,但如何训练优化模板是影响该方法提取精度的关键所在。笔者引入粒子群优化算法进行参数优化,结合细胞神经网络和粒子群方法各自的特点,通过细胞神经网络动态性能分析,对模板加入约束条件,并在粒子群算法中引入收缩因子,提出一种基于改进的粒子群优化的细胞神经网络算法。使用该方法进行了模型验证和实际资料处理,处理结果表明,该方法提高了重力异常提取的稳定性,能够更准确地提取油气藏重力异常信息。     

神经智能信息处理系统的研究现状及其在水文水资源中的应用展望

从神经网络(NN),模糊系统(FS)和进化算法(EC)三者相结合的角度,介绍了神经智能信息处理系统的研究现状,讨论了该系统在水文水资源中的应用,并对其应用发展趋势进行了预测。     

边坡稳定性预测的模糊神经网络模型

根据边坡稳定问题具有的模糊性,提出了一种判定边坡稳定性的模糊神经网络模型。该系统仅从期望输入输出数据集即可达到获取知识、确定模糊初始规则基的目的。再利用神经网络学习能力便不难修改规则库中的模糊规则以及隶属函数和网络权值等参数,这样大大减少了规则匹配过程,加快了推理速度,从而极大程度地提高了系统的自适应能力。最后用收集到的边坡数据样本训练和测试模糊神经网络模型,结果表明该模糊神经网络预测边坡稳定性是可行的、有效的。     

基于模糊神经网络的油气综合评判方法研究

针对常规模式识别方法进行油气预测时,存在样本数量大、参数非线性及已知样本数类型分类有明显的模糊性等特点。本文根据神经网络的自适应谐振理论结合模糊Ｋ均值聚类算法、特征参数非线性映射的有效性校验,构成了一个性能完善的地震参数油气预测系统。该系统对样本特征空间用模糊神经网络建立未知样本的预测度量,更加符合客观地质特征和油气分布规律,从而能判别油气富集的有利区域。数值分析和实践应用表明,该方法数值稳定可靠     

基于模糊神经网的油气综合评判方法研究

针对常规模式识别方法进行油气预测时,存在样本数量大、参数非线性及已知样本数类型分类有明显的模糊性等特点。本文根据神经网络的自适应谐振理论（ＡｄａｐｔｉｖｅＲｅｓｏｎａｎｃｅＴｈｅｏｒｙ）结合模糊Ｋ均值聚类算法、特征参数非线性映射的有效性校验,构成了一个性能完善的地震参数油气预测系统。该系统对样本特征空间用模糊神经网络建立未知样本的预测度量,更加符合客观地质特征和油气分布规律,从而能判别油气富集的有利区域。数值分析和实践应用表明,该方法数值稳定可靠,在储层油气预测方面有较好的效果     

新场气田开发方案寻优的遗传优化神经网络模型

在深入剖析遗传算法GA和BP神经网络优势的基础上,提出了改进的遗传优化BP神经网络模型(简称GABP模型).该原理是利用遗传算法(GA)对BP神经网络的结构和连接权进行全局优化学习,从而获取最佳的网络模型.经前人大量研究表明:遗传算法的适应性好,对搜索空间没有任何特殊要求,且全局优化能力强,可以有效地在整个解空间寻优,但遗传算法存在局部调节能力较弱、容错性较差的不足.而神经网络的容错性较强,具有自组织、自适应和分布式储存的特点,它可以通过学习乖训练,进行模型结构的自组织,适应不同信息或信息模糊,以及推理规则不明确问题的处理,但神经网络方法又有搜索能力较差,容易陷入局部解之中的缺点.从上述二种方法的优缺点出发,按照取其所长,克己所短的原则,将二种方法有机地结合起来,构建了GABP模型及其算法.通过对新场气田开发方案优选的决策应用,表明该模型的评价结果准确合理,可为类似研究借鉴参考.     

神经网络串行训练并行处理

作者在文中提出一种神经网络串行训练处理方法，它能提高神经网络学习收敛的速度，并达到提高训练精度的要求，此外，该算法还可依据实际问题的需求，自适应地扩展神经网络的容量，使得可最大限度地充分利用已知信息，达到增强神经网络预测结果可靠性的能力，作者在文中还给出了本文算法运用于大庆油田甘开发区块储层油气预测应用实例，效果良好。     

The terrestrial fauna of the Late Triassic Pant-y-ffynnon Quarry fissures,South Wales,UK and a new species of Clevosaurus (Lepidosauria: Rhynchocephalia)

Pant-y-ffynnon Quarry in South Wales yielded a rich cache of fossils in the early 1950s, including articulated specimens of new species (the small sauropodomorph dinosaur Pantydraco caducus and the crocodylomorph Terrestrisuchus gracilis), but no substantial study of the wider fauna of the Pant-y-ffynnon fissure systems has been published. Here, our overview of existing specimens, a few described but mostly undescribed, as well as freshly processed material, provides a comprehensive picture of the Pant-y-ffynnon palaeo-island of the Late Triassic. This was an island with a relatively impoverished fauna dominated by small clevosaurs (rhynchocephalians), including a new species, Clevosaurus cambrica, described here from a partially articulated specimen and isolated bones. The new species has a dental morphology that is intermediate between the Late Triassic Clevosaurus hudsoni, from Cromhall Quarry to the east, and the younger C. convallis from Pant Quarry to the west, suggesting adaptive radiation of clevosaurs in the palaeo-archipelago. The larger reptiles on the palaeo-island do not exceed 1.5?m in length, including a small carnivorous crocodylomorph, Terrestrisuchus, and a possible example of insular dwarfism in the basal dinosaur Pantydraco.     

The effect of water on accessory phase solubility in subaluminous and peralkaline granitic melts

The solubilities of columbite, tantalite, wolframite, rutile, zircon and hafnon were determined as a function of the water contents in peralkaline and subaluminous granite melts. All experiments were conducted at 1035 °C and 2 kbar and the water contents of the melts ranged from nominally dry to approximately 6 wt.% H₂O. Accessory phase solubilities are not affected by the water content of the peralkaline melt. By contrast, solubilities are affected by the water content of the subaluminous melt, where the solubilities of all the accessory phases examined increase with the water content of the melt, up to 2 wt.% H₂O. At higher water contents, solubilities are nearly constant. It can be concluded that water is not an important control of accessory phase solubility, although the water content will affect diffusivities of components in the melt, thus whether or not accessory phases will be present as restite material. The solubility behaviour in the subaluminous and peralkaline melts supports previous spectroscopic studies, which have observed differences in the coordination of high field strength elements in dry vs. wet subaluminous granitic glasses, but not for peralkaline granitic glasses. Lastly, the fact that wolframite solubility increases with increasing water content in the subaluminous melt suggests that tungsten dissolved as a hexavalent species.     

Quaternary stresses revealed by calcite twinning inversion: insights from observations in the Savonnières underground quarry (eastern France)

Calcite samples were extracted both from the rock matrix and the superficial coating of a karstified fault plane of an underground quarry, located in the eastern border of the Paris basin. The karstification is dated as Quaternary. Analysis of mechanical calcite twinning reveals that only the calcite matrix has also undergone a compression trending WNW that can be attributed to the Mio-Pliocene alpine collision. Both coating and matrix have undergone a strike-slip regime with σ₁ roughly trending north–south, that could correspond to the regional present-day state of stress, a strike-slip compression rather trending NNW, modified by local phenomena. To cite this article: M. Rocher et al., C. R. Geoscience 335 (2003).     

HYDROGEOLOGY

正20141756 Chen Ruige(Mathematical College,China University of Geosciences,Beijing100083,China);Zhou Xun Numerical Simulation of Groundwater Level Fluctuation in a Coastal Confined Aquifer with Sloping Initial Groundwater Level Induced by the Tide(Geological Bulletin of China,ISSN1671-2552,CN11-4648/P,32(7),2013,p.1099-1104,6 illus.,16 refs.) Key words:confined water,groundwater level     

METAMORPHIC PETROLOGY

正20141408 Cai Jia(Institute of Geology,Chinese Academy of Geological Sciences,Beijing100037,China);Liu Fulai Petrogenesis and Metamorphic P-T Conditions of Garnet-Spinel-Biotitebearing Paragneiss in Danangou Area,Daqingshan-Wulashan Metamorphic Complex Belt(Acta Petrologica Sinica,ISSN1000-0569,CN11-1922/P,29(7),     

SERIALS

正~~     

GEOPHYSICAL EXPLORATION

正20142386An Guoying(China Aero Geophysical Survey and Remote Sensing Center for Land and Resources,Beijing 100083,China)Application of Satellite Remote Sensing in Regional Hydrogeological Investigation:Taking Cenozoic Strata in Wenquan Sheet(1∶250 000)of Karakoram Range as an Example(Geosci-     

QUATERNARY GEOLOGY & GEOMORPHOLOGY

正20141016An Chengbang(Key Laboratory of Western China’s Environmental Systems,Ministry of Education,Lanzhou University,Lanzhou 730000,China);Zhao Yongtao Lake Records during the Last Glacial Maximum from Xinjiang,NW China and Their Climatic Impli-     

PETROLEUM GEOLOGY

正20141538 Cao Qing(School of Earth Sciences and Engineering,Xi’an Petroleum University,Xi’an 710065,China);Zhao Jingzhou Characteristics and Significance of Fluid Inclusions from Majiagou Formation,Yichuan Huangling Area,Ordos Basin(Advances in Earth Science,ISSN1001-8166,CN62-1091/P,28(7),2013,p.819-828,7 illus.,3 tables,43 refs.)     

GEOCHEMISTRY

正20142002 Wei Hualing(Institute of Geophysical and Geochemical Exploration,Chinese Academy of Geological Sciences,Langfang065000,China);Zhou Guohua Element Content and Mineral Compositions in Different Sizes of Soil in Tongling Area,Anhui Province(Geological Bulletin of China,ISSN1671-2552,CN11-4648/P,32(11),2013,p.1861