基于GA-SVR的渗透系数参数反演方法          下载免费PDF全文

陈海洋  膝彦国  王金生 《水文地质工程地质》2011,38(2):14-18

渗透系数参数反演的本质是优化问题求解,遗传算法是一种基于自然选择和群体遗传机理的新的全局优化求解方法,可以较好地用于求解诸如渗透系数参数反演等复杂非线性组合优化问题。基于结构风险最小化原理的支持向量机具有逼近复杂非线性系统、较强的学习泛化能力,可以用来计算渗透系数参数反演过程中的测点水头值。实验表明,基于遗传算法-支持向量回归机的地下水渗透系统参数反演拟合效果良好,能大大提升区间搜索效率,避免出现局部最优解,其参数识别精度符合实际应用要求。  相似文献   

含水层对地下水热泵承载力数值评价及其实用化:以北京平原区为例   总被引：1，自引：1，他引：0  

李世君  刘立才  郑凡东  江剑 《现代地质》2011,25(2):370-376

针对北京平原区的水文地质条件,运用地下热传递数值法评价了年内地下热量均衡、失衡两种状态下,各水文地质亚区单位面积含水层对地下水热泵的承载力,并利用多元线性回归方法阐明了承载力、抽灌井间距与含水层各参数之间的关系。研究表明:承载力仅与含水层厚度显著正相关,而抽灌井间距与含水层的渗透系数和厚度均显著正相关。研究结果为类似地区的地下水热泵定量规划提供了简捷的估算方法及参考依据。  相似文献   

Comparison of Two Sample Preparation Methods for X‐Ray Fluorescence Spectrometry in the Determination of Ni and Cr     

Surendra P. Verma  Rosalinda González‐Ramírez  Rodolfo Rodríguez‐Ríos 《Geostandards and Geoanalytical Research》2011,35(2):183-192

Pressed powder pellets and fused beads or glass disks are routinely used in X‐ray fluorescence spectrometry for the determination of major and trace elements, respectively, in geological materials. In order to evaluate the performance of these two sample preparation methods, we determined Ni and Cr concentrations of fourteen RMs from Japan, France and South Africa, and eighty‐five igneous and three sedimentary rock samples from Mexico in both powder pellets and glass beads. We also computed new values of statistical parameters for RMs from an outlier‐based multiple‐test method and compared them with the literature mean and confidence limit values. The results showed that the multiple‐test method provided more reliable central tendency and dispersion parameters for RMs than those obtained previously from the two or three standard deviation method, or from robust methods. The powder pellet and fused bead sample preparation methods provided consistent results for Ni and Cr at concentration levels > 50 μg g^?1 in this application; for lower concentration levels, however, these methods showed somewhat greater differences. For quantitative comparisons, both ordinary and weighted least‐squares linear regression models were used to show that the two sample preparation methods provided generally unbiased results.  相似文献   

吉林西部土壤有机碳含量与元素地球化学特征     

王冬艳  杨园园  李月芬  王孟  郭冬艳  靳克 《吉林大学学报(地球科学版)》2011,(Z1):333-338

针对土壤有机碳影响因素多且复杂而造成碳储量模拟模型中参数难以确定的问题,利用吉林西部167个典型土壤样点资料,通过统计分析对吉林西部表层土壤有机碳含量及元素地球化学进行了详细研究。结果表明:吉林西部土壤有机碳含量整体水平偏低,并存在着明显的区域分布特点,不同工作区的分异性不同;吉林西部土壤有机碳含量与Si含量、硅铝率、硅铝铁率和K含量之间存在着极显著的负相关,而与P、Al、Fe、Ca、Mg含量存在极显著的正相关;不同元素变异特征值可以解释土壤有机碳含量总变异的能力也不相同。通过研究土壤有机碳与元素含量的关系来定量研究土壤有机碳与土壤理化性质的关系是可行的,将二者量化关系与土壤有机碳储量模拟模型相结合,使得对土壤碳储量的认识更加科学。  相似文献   

Prediction of geometrical properties of perfect breaking waves on composite breakwaters     

Mevlut Sami Akoz  M. Salih Kirkgoz 《Applied Ocean Research》2011,33(3):178-185

Breaking wave loads on coastal structures depend primarily on the type of wave breaking at the instant of impact. When a wave breaks on a vertical wall with an almost vertical front face called the “perfect breaking”, the greatest impact forces are produced. The correct prediction of impact forces from perfect breaking of waves on seawalls and breakwaters is closely dependent on the accurate determination of their configurations at breaking. The present study is concerned with the determination of the geometrical properties of perfect breaking waves on composite-type breakwaters by employing artificial neural networks. Using a set of laboratory data, the breaker crest height, h_b, breaker height, H_b, and water depth in front of the wall, d_w, from perfect breaking of waves on composite breakwaters are predicted using the artificial neural network technique and the results are compared with those obtained from linear and multi-linear regression models. The comparisons of the predicted results from the present models with measured data show that the h_b, H_b and d_w values, which represent the geometry of waves breaking directly on composite breakwaters, can be predicted more accurately by artificial neural networks compared to linear and multi-linear regressions.  相似文献   

Architecture of a forced regressive systems tract in the Turonian Ferron “Notom Delta”, southern Utah, U.S.A.     

Weiguo Li  Janok BhattacharyaYijie Zhu 《Marine and Petroleum Geology》2011,28(8):1517-1529

The large-scale stratigraphic architecture of forced regressive deposits has been documented in many previous studies. Bed-scale facies architectural analyses of these deposits, however, are still very limited. The Cretaceous Ferron “Notom Delta” in southern Utah, U.S.A. contains a 20 km dip-oriented exposure of a stepped, forced regressive systems tract. The main focus of this paper is to reconstruct the paleogeography and depositional history of the systems tract based on detailed stratigraphic and facies architectural analysis using 23 geological sections, photomosaics, and walking out of beds.Internally, the systems tract consists of 6 parasequences, 11f to 11a from the oldest to the youngest. During the progradation of parasequences 11f to 11b the paleoshorelines were wave-dominated, as indicated by the abundance of HCS and/or SCS beds, wave-ripple cross-laminated beds, and the occurrence of diverse and robust ichnological suites attributable to the Skolithos and Cruziana Ichnofacies. Progradation of the wave-dominated shorelines resulted in more homogeneous and laterally continuous sand bodies. From 11b to 11a, however, there is a distinct change in paleoshoreline regime from wave-dominated to tide-influenced as indicated by the common occurrence of tidal facies in 11a, including: (1) lenticular, wavy, and flaser bedding and bidirectional dipping cross strata; (2) reactivation surfaces, double-mud drapes, and ripple cross lamination with opposing dips at the toe of large dune-scale cross sets; (3) inclined heterolithic strata (IHS) and sigmoidal bedding with tidal rhythmites; and (4) cyclic vertical variation in facies and bed thickness and the common occurrence of sand-mud couplets. These tide-influenced facies show overall lower bioturbation intensity (BI 0-3). Progradation of the tide-influenced shoreline results in more heterolithic delta-front facies. Tidal and/or tidal-fluvial channels further dissect delta-front sandstones, forming more isolated sand bodies.Data from this study, as well as previous work, show that width and thickness of the forced regressive parasequences are small, typically less than 5 km and 20 m respectively. In subsurface studies, identifying and correlating such small-scale parasequences using sparse data involve significant uncertainties. A combination of the diagnostic features indicating forced regression and different data sets is essential to better constrain the geometry and architecture these small-scale bodies.  相似文献   

基于机器学习的复杂地形下短期数值天气预报误差分析与订正     

任萍  陈明轩  曹伟华  王在文  韩雷  宋林烨  杨璐 《气象学报》2020,78(6):1002-1020

初步研发了一套基于机器学习方法XGBoost且考虑地形特征影响的数值预报多模式集成技术,并与传统的等权重平均和线性回归方法的集成效果进行了对比分析。利用北京地区快速更新循环数值预报系统每天8次循环预报给出的近地面2 m温度、2 m相对湿度、10 m风速、10 m风向数据产品,分别基于机器学习方法XGBoost、等权重平均方法、线性回归方法构建了3种体现地形因子影响的多模式预报时间滞后集成模型。试验对比分析了暖季、冷季每日不同时刻的模式预报集成订正效果。结果表明:分季节试验中,基于XGBoost模型对2 m温度、10 m风速的集成预报结果相对原始最优预报结果误差明显优于其他两种传统方法。XGBoost对2 m温度集成的误差可降低11.02%—18.09%,10 m风速集成误差可降低31.23%—33.22%,10 m风向集成误差可降低4.1%—8.23%。2 m相对湿度的集成预报误差与传统方法接近。基于XGBoost的多模式集成预报模型可以充分“挖掘”不同模式或不同时刻快速更新循环预报优点,有效降低模式的系统性误差,提供准确性更高的多模式集成确定性预报产品。   相似文献   

基于空间回归的B样条拟合在地面气温资料质量控制中的应用     

熊雄  姚薇  叶小岭  张颖超  杨帅 《热带气象学报》2020,36(4):489-498

将B样条拟合算法引入到地面气温资料的质量控制当中,在分析地面气温资料空间相关性的基础上,提出一种基于空间回归的B样条拟合地面气温资料质量控制方法(SRT_BSF方法)。为了检验SRT_BSF方法的有效性及适应性,利用SRT_BSF方法对多个场景地面气温资料进行质量控制,并与反距离加权方法(IDW方法)和空间回归方法(SRT方法)进行比较分析。试验结果表明,SRT_BSF方法相对于IDW方法和SRT方法更能有效地标记出地面气温资料中的存疑数据,同时多组独立案例的分析结果说明SRT_BSF方法具有更好的稳定性和适用性。   相似文献   

动态时变参数方法在云量精细化预报中的应用研究     

杜晖  殷启元  贾晓红  姬雪帅  尚可政  尚子溦  曾瑛 《热带气象学报》2020,36(6):834-845

利用2008年1月—2013年12月以及2017年1—11月全球天气预报系统（GFS）预报场资料,采用自适应线性最小二乘回归（LS）和自适应递推卡尔曼（Kalman）滤波两种动态时变参数方法,建立了河套周边地区0~168 h预报时效的总云量精细化预报,并与GFS模式直接输出的总云量、线性预报模型逐步回归预报方法得到的总云量以及非线性预报模型BP神经网络和最小二乘支持向量机回归方法（LSSVM）得到的总云量进行了对比,结果如下:（1）相比GFS模式直接输出的总云量,LS、BP神经网络、LSSVM得到的总云量与实况值的平均绝对误差均明显减小。LS方法误差最小,LS方法的年MAE均在20%~25%,且随着预报时效的延长,改进效果越大。LS方法、多元逐步回归方法、BP神经网络、LSSVM四种方法在6—8月的改进效果最大。（2）LS方法预报的总云量与实况云量的相关性最好,即使168 h预报时效的相关系数依然在0.64以上,远高于其他几种模型的预报结果。（3）LS方法能够明显地提高少云和多云天空状况下预报的击中率,且最优（少云击中率平均提高24 %,多云击中率平均提高34 %）。（4）自适应递推Kalman滤波方法存在预报滞后现象,改进效果不明显。   相似文献   

基于多种模型的云南元江哈尼云海景观预报研究     

丁圣  段玮  朱勇  李刚 《气象与环境学报》2020,36(3):106-112

为推动区域旅游事业发展,满足旅游气象服务需求,研究针对云南元江哈尼云海景观进行了立体气象观测和业务预报实验。本研究基于云南元江云海气候站2016—2019年观测数据,利用经验预报法、Logistics回归、支持向量机、决策树分析等方法,进行了云海景观出现与否的二分类预报实验。结果表明：各预报方法间训练样本总体准确率在74.3%—82.2%之间差别不大,但传统经验预报基于云海机理研究背景,预报指标物理意义明确,随着预报经验的积累经验预报2019年TS评分为54.8,优于2016—2018年TS评分46.0,也优于仅使用局地数据的统计学习算法的预报评分,且其他几种统计学习预报方法的检验样本TS评分均不如训练样本评分高。云海景观出现需要水汽条件和大气静稳条件的配合,局地云海气象观测站建设收集的立体气候数据有利于预报人员改进预报指标体系,提高预报准确率,有利于提升区域旅游气象服务能力发展。  相似文献