四川旱涝震灾害的人工神经网络外推预测

选取旱、涝、震灾害之间的一步转移概率作为Ｂ－Ｐ人工神经网络训练样本的输入信息，建立了四川旱、涝、震的人工神经网络灾型预测模型。Ｂ－Ｐ网络模型应用于实例预测结果与用主分量分析法的趋势预测结果精度接近。     

基于数据分布域变换与贝叶斯神经网络的渗透率预测及不确定性估计

渗透率是储层评价和油气藏开发的关键参数.传统测井方法与常规机器学习方法估算的渗透率都是固定值.但由于测井数据本身存在噪声, 渗透率的预测结果可能受到噪声的影响出现测量性的随机误差(即任意不确定性); 同时, 当测试数据与训练数据存在差异时, 机器学习模型在预测渗透率时可能出现模型参数的不确定性(即认知不确定性).为实现渗透率的准确预测并量化两种不确定性对结果的影响, 本文提出基于数据分布域变换和贝叶斯神经网络同时实现渗透率预测及其不确定性的估计.提出方法主要包括两个部分: 一部分是不同域数据分布的相互转换, 另一部分是基于贝叶斯理论的神经网络渗透率建模预测和不确定性估计.由于贝叶斯神经网络存在数据分布的假设, 当标签的概率分布与网络的分布保持一致时, 贝叶斯神经网络可以更好的学习到数据之间的关系.因此通过寻找一个函数将一个原始域的渗透率标签转换为目标域的与渗透率有关的变量(我们称为目标域渗透率), 使得该变量符合贝叶斯神经网络的分布假设.我们使用贝叶斯神经网络预测目标域渗透率以及任意不确定性和认知不确定性.随后, 通过分布域的逆变换, 我们将目标域渗透率还原回原始域渗透率.应用本文方法到某油田的18口井的测井数据中, 使用16口井的数据进行训练, 2口井进行测试.测试井的预测渗透率与真实渗透率基本一致.同时, 任意不确定性的预测结果提供了渗透率预测值受到的测井数据噪声影响的位置.认知不确定的预测结果说明数据量少的位置具有更高的认知不确定性.我们提出的这一流程不仅显示了在储层表征方面的巨大潜力, 同时可以降低测井解释时的风险.

     

Beach nourishments at Coolangatta Bay over the period 1987–2005: Impacts and lessons

Erosion of the southern Gold Coast beaches (SE Queensland, Australia) was exacerbated after the extension of the Tweed River training walls in the early 1960s. To achieve the objective of restoring and maintaining beach amenity, significant nourishment works have been undertaken in Coolangatta Bay over the past 30 years. Particularly, under the Tweed River Entrance Sand Bypassing Project (TRESBP) since 1995, a number of nourishment campaigns and the implementation of a permanent sand bypass system in 2001 have resulted in significant changes of Coolangatta Bay morphology. The present case study investigates the influence of both wave climate and nourishment works on the area extending from the updrift Snapper Rocks area to downdrift Kirra Beach. SWAN spectral wave model is implemented at Coolangatta Bay area and forced by the global wave model WW3 to estimate wave forcing and the potential natural longshore drift entering in Coolangatta. Specific transects extracted from accurate bathymetric surveys are used to investigate and quantify Coolangatta Bay sedimentation for the period 1987–2005. A network of Argus video stations provides high sample rate information on the shoreline evolution. Results show that, over the past 10 years, Coolangatta Bay has infilled rapidly. Sedimentation reached up to 6 m in some areas between 1995 and 2005, with beach width increasing by 200 m at Kirra Beach. Rapid seaward shoreline migration is consistent with the intense over-pumping of sand relative to the natural potential to move sand alongshore. The nourishment strategy used during this project has successfully delivered large amounts of sand to the southern Gold Coast embayment, although it has been up to now controversial from many community perspectives. The artificial sand bypassing process proved to be much more efficient than depositing the dredged sand in the nearshore area which requires a significant period of low energy condition in order for the deposited sediment to migrate shoreward and weld to the shore. This case study confirms that, when carefully undertaken, sand bypassing is a sustainable and flexible soft engineering approach which can work in concert with natural processes.     

变形岩石磁性各向异性及其应变相关性的模拟

本文介绍处于弱磁场中的磁性砂岩人工样品,在三种应变作用下磁化率各向异性特征的研究结果.结果表明,磁化率椭球与应变椭球之间,显示了密切的相关性.文中并对上述情况下,磁化率各向异性特征及剩磁特征,从机理上进行了分析.     

地震油气储层的小样本卷积神经网络学习与预测

地震储层预测是油气勘探的重要组成部分,但完成该项工作往往需要经历多个环节,而多工序或长周期的研究分析降低了勘探效率.基于油气藏分布规律及其在地震响应上所具有的特点,本文引入卷积神经网络深度学习方法,用于智能提取、分类并识别地震油气特征.卷积神经网络所具有的强适用性、强泛化能力,使之可以在小样本条件下,对未解释地震数据体进行全局优化提取特征并加以分类,即利用有限的已知含油气井段信息构建卷积核,以地震数据为驱动,借助卷积神经网络提取、识别蕴藏其中的地震油气特征.将本方案应用于模型数据及实际数据的验算,取得了预期效果.通过与实际钻井信息及基于多波地震数据机器学习所预测结果对比,本方案利用实际数据所演算结果与实际情况有较高的吻合度.表明本方案具有一定的可行性,为缩短相关环节的周期提供了一种新的途径.

     

甘宁青地区强震孕育的深部电性和力学环境

以200多点次的大地电磁测深资料和6条人工地震测深剖面的资料为根据，结合重力测量和地热流观测资料，分析研究了甘肃、宁夏和青海地区的强震孕育的深部环境。认为强震孕育在增厚的低阻层的顶部，低速层的上方以及介质在横向上电性和力学性质均剧烈变化的环境之中。这些条件的实质是在地应力作用下纵、横向地层变形不一致，在这不协调之处极有利于应力集中。     

Classification of seismic signals at Villarrica volcano (Chile) using neural networks and genetic algorithms

Each volcano has its own unique seismic activity. The aim of this work is to construct a system able to classify seismic signals for the Villarrica volcano, one of the most active volcanoes in South America. Since seismic signals are the result of particular processes inside the volcano's structure, they can be used to forecast volcanic activity. This paper describes the different kinds of seismic signals recorded at the Villarrica volcano and their significance. Three kind of signals were considered as most representative of this volcano's activity: the long-period, the tremor, and the energetic tremor signals. A classifier is implemented to read the seismic registers at 30-second intervals, extract the most relevant features of each interval, and classify them into one of the three kinds of signals considered as most representative of this particular volcano. To do so, 1033 different kinds of 30-s signals were extracted and classified by a human expert. A feature extraction process was applied to obtain the main characteristics of each of them. This process was developed using criteria which have been shown by others to effectively classify seismic signals, based on the experience of a human expert. The classifier was implemented with a Multi-Layer Perceptron (MLP) artificial neural network whose architecture and training process were optimized by means of a genetic algorithm. This technique searched for the most adequate MLP configuration to improve the classification performance, optimizing the number of hidden neurons, the transfer functions of the neurons, and the training algorithm. The optimization process also performed a feature selection to reduce the number of signal features, optimizing the number of network inputs. The results show that the optimized classifier reaches more than 93% exactitude. identifying the signals of each kind. The amplitude of the signals is the most important feature for its classification, followed by its frequency content. The described methodology can be used to classify more seismic signals to improve the study of the activity of this volcano or to extend the study to other active volcanoes of the region.     

我国大黄鱼养殖产业发展研究

文章对我国大黄鱼养殖产量及分布、养殖模式、投入产出情况、收购价格变动和进出口情况进行分析,总结了几项产业中存在的问题,包括：养殖结构亟待调整、关键养殖技术欠缺、鱼苗成活率偏低、市场中存在逆向选择。在此基础上,提出积极调整养殖结构与养殖密度、提升合作社与试验站职能、加强养殖技术与鱼药研发力度、培育推广优质大黄鱼苗种、促进可追溯体系建设的产业发展建议。     

某异型钢管混凝土拱桥地震反应分析

本文以某异型钢管混凝土拱桥为研究对象,通过ANSYS建立了结构的空间有限元模型,计算和分析了该桥的动力特性。同时,通过MATLAB程序生成拟合规范反应谱的人工地震波,并运用时程分析法计算了该桥在一维和多维输入下的地震反应,分析了该桥的地震反应规律,为钢管混凝土拱桥的抗震性能分析提供了一定的依据。     

Irrigated area determination: a case study in the Province of San Juan,Argentina

Much of the central-western region of Argentina, where San Juan Province is located, experiences arid to semi-arid climatic conditions with low average annual rainfall accompanied by substantial evapotranspiration. Consequently, a viable crop industry depends to a large extent upon irrigation from major river systems. Increasing demand for water in the lower basin of the San Juan River is emphasizing the need for more accurate estimates of water used for irrigation. Since the water demand for a particular crop is very closely related to crop area, monitoring the area of crop under irrigation is considered a proxy for the amount of water used. Landsat 5 imagery for the growing season, field data and aerial photographs were used to evaluate crop area.