A Real-time Lithological Identification Method based on SMOTE-Tomek and ICSA Optimization

In petroleum engineering, real-time lithology identification is very important for reservoir evaluation, drilling decisions and petroleum geological exploration. A lithology identification method while drilling based on machine learning and mud logging data is studied in this paper. This method can effectively utilize downhole parameters collected in real-time during drilling, to identify lithology in real-time and provide a reference for optimization of drilling parameters. Given the imbalance of lithology samples, the synthetic minority over-sampling technique (SMOTE) and Tomek link were used to balance the sample number of five lithologies. Meanwhile, this paper introduces Tent map, random opposition-based learning and dynamic perceived probability to the original crow search algorithm (CSA), and establishes an improved crow search algorithm (ICSA). In this paper, ICSA is used to optimize the hyperparameter combination of random forest (RF), extremely random trees (ET), extreme gradient boosting (XGB), and light gradient boosting machine (LGBM) models. In addition, this study combines the recognition advantages of the four models. The accuracy of lithology identification by the weighted average probability model reaches 0.877. The study of this paper realizes high-precision real-time lithology identification method, which can provide lithology reference for the drilling process.     

Teaching Structural Hazards Awareness for Preparedness and Community Response

Most successful search and rescue is performed by neighbors and household members during the first 24 h after an earthquake, before professional responders arrive on the scene. Many spontaneous responders become casualties in the process. Kandilli Observatory and Earthquake Research Institutes Disaster Preparedness Education Project with the support of the Earthquake Engineering Department has developed a training module for Community Disaster Responders to help them assess the level of damage for immediate citizen disaster response. Learning the signs of damage in both reinforced concrete and adobe buildings, they will be better able to distinguish the slight and moderately damaged buildings where their light search and rescue efforts will be effective, as opposed to the heavily damaged buildings where professional help is required for safe extrication. Incorporating citizens into the early stages of building collapse rescue is designed to make professional urban search and rescue efforts more effective. The program is being piloted in Turkey by the Civil Defense Directorate, Search and Rescue Brigade. It is expected that this will lead to a new demand for engineers to work in partnership with community-based organizations for planning and prioritizing disaster mitigation efforts, and to participate in immediate post-disaster damage assessment. Engineering recommendations will also be needed for post-earthquake building re-entry and re-habitation risk assessment.     

供水管网管理系统的设计与实现

城市管网系统的管理是城建工作的一个重要课题。本文介绍已开发实现的青岛市《供水管网管理系统》的设计及主要的技术特点     

On the existence of small comets and their interactions with planets

Arguments are presented for a substantial, unexplored population of comets with radii less than 1 km. Known examples confirm this population and extrapolation of any plausible size-distribution function indicates large numbers. However, their accurate numbers, orbital characteristics, and physical properties are unknown. Thus, even though the small comets may be the most frequent cometary bodies impacting the planets, a quantitative evaluation is not currently possible. We advocate an optimized, dedicated search program to characterize this population.Laboratory for Atmospheric and Space Physics, University of Colorado at BoulderLaboratory for Atmospheric and Space Physics, University of Colorado at BoulderLaboratory for Atmospheric and Space Physics, University of Colorado at Boulder     

Target intersection probabilities for parallel-line and continuous-grid types of search

The expressions for calculating the probability of intersection of hidden targets of different sizes and shapes for parallel-line and continuous-grid types of search can be formulated by vsing the concept of conditional probability. When the prior probability of the orientation of a widden target is represented by a uniform distribution, the calculated posterior probabilities are identical with the results obtained by the classic methods of probability. For hidden targets of different sizes and shapes, the following generalizations about the probability of intersection can be made: (1) to a first approximation, the probability of intersection of a hidden target is proportional to the ratio of the greatest dimension of the target (viewed in plane projection) to the minimum line spacing of the search pattern; (2) the shape of the hidden target does not greatly affect the probability of the intersection when the largest dimension of the target is small relative to the minimum spacing of the search pattern, (3) the probability of intersecting a target twice for a particular type of search can be used as a lower bound if there is an element of uncertainty of detection for a particular type of tool; (4) the geometry of the search pattern becomes more critical when the largest dimension of the target equals or exceeds the minimum spacing of the search pattern; (5) for elongate targets, the probability of intersection is greater for parallel-line search than for an equivalent continuous square-grid search when the largest dimension of the target is less than the minimum spacing of the search pattern, whereas the opposite is true when the largest dimension exceeds the minimum spacing; (6) the probability of intersection for nonorthogonal continuous-grid search patterns is not greatly different from the probability of intersection for the equivalent orthogonal continuous-grid pattern when the orientation of the target is unknown. The probability of intersection for an elliptically shaped target can be approximated by treating the ellipse as intermediate between a circle and a line. A search conducted along a continuous rectangular grid can be represented as intermediate between a search along parallel lines and along a continuous square grid. On this basis, an upper and lower bound for the probability of intersection of an elliptically shaped target for a continuous rectangular grid can be calculated. Charts have been constructed that permit the values for these probabilities to be obtained graphically. The use of conditional probability allows the explorationist greater flexibility in considering alternate search strategies for locating hidden targets.This paper was presented at Symposium 116.3, Quantitative Strategy for Exploration, held as part of the 25th International Geological Congress, Sydney, Australia, August 1976.     

The budgeting problem of a searching firm

This paper explores factors to be taken into consideration in determining the optimal exploration budget for a mining company readying itself for detail work within a mineral province. Optimal strategy is shown to be a combination of results from recent conventional reservation value stopping rules and optimal expenditure models appearing earlier in the literature.     

中国钾盐地质资源现状与找钾方向初步分析

分析了中国钾资源供需形势、有利成盐找钾时代、主要的有利成钾盆地等,回顾了中国找钾工作的历史,指出中国西部仍有找钾前景,尤其是中、上侏罗统-白垩系和新生界,其次是上震旦统-寒武系、石炭系、奥陶系和三叠系,同时,还应继续开展现代盐湖和地下卤水找钾和综合评价,主要找钾地区为塔里木盆地、滇西-羌塘盆地、柴达木盆地和上扬子盆地.鉴于中国特定的成盐构造地质条件和钾盐成矿、保存特点,以及国外预测找钾的经验,在中国找钾要有做长期深入勘查研究的思想准备和采取切实有效的措施,建议对找钾工作加大投入,加强找钾预测的地质基础研究,建立专业队伍和落实油钾兼探的方针.     

面向应用的高光谱影像分类方法

提出了一种面向应用的高光谱影像分类方法,旨在从根本上、全方位地削弱各种不利因素对该类影像分类精度的影响。主要包括利用IEM算法获取更为精确的类别分布信息,采用Tabu搜索算法进行原始特征空间的降维,运用基于混合规则的组合分类器来判断待识样本的类别标签。实验表明,按照该方法进行高光谱影像的分类处理,可以得到很高精度的分类结果。     

输出误差系统的多新息辨识方法

研究了输出误差(OE)系统和输出误差自回归滑动平均(OEARMA)系统(即Box-Jenkins系统)的辅助模型随机梯度算法、辅助模型多新息随机梯度算法、辅助模型递推最小二乘算法、辅助模型多新息最小二乘算法,输出误差系统的修正辅助模型随机梯度算法、遗忘因子辅助模型随机梯度算法、变递推间隔辅助模型随机梯度算法、变递推间隔辅助模型多新息随机梯度算法、变递推间隔辅助模型递推最小二乘算法等,以及输出误差自回归(OEAR)系统的基于滤波的辅助模型多新息广义随机梯度算法和基于滤波的辅助模型多新息广义最小二乘算法.     

主动源海底地震仪横波模型的自动搜索与非唯一性分析*

主动源海底地震仪探测在海底结构的研究中发挥着重要作用, 其中转换横波数据模拟为研究海底构造和物质属性提供了精确依据。本文针对现行转换横波模拟技术存在的步骤繁琐、难以确保最优解和无法进行非唯一性分析等问题进行研究, 提出了基于模型解空间和目标函数的模拟技术, 形成了主动源转换横波数据模拟的新方法, 该方法可借助计算机程序实现结构模拟的自动化。在南海西北陆缘的西沙地块OBS2013-3测线上对该方法进行验证, 分别利用单台PPS震相和全体台站的PSS震相走时数据进行模拟试验。结果表明, 本文的方法能够提供对于最优模型的快速、准确搜索和非唯一性范围的估计。这一方法有助于提高主动源海底地震仪转换横波数据模拟的效率, 并为结果的可靠性和稳定性提供更好的保障。