通用模型参数率定技术研究

在介绍当前模型参数优选技术的基础上，设计了模型参数同优选方法相耦合的一系列数据接口定义，实现了人工试错和自动优选相耦合、多模型多参数同时自动优选的模型参数率定功能。应用实例表明，通用模型参数率定功能具有简便、快捷、准确的优点，大大提高了参数率定的效率。     

新安江模型参数多目标优化研究

水文模型的参数优化率定一直以来是水文预报领域的重要研究内容,当水文模型的结构确定后,水文模型参数的选择对水文模型整体性能和水文预报结果的好坏有着至关重要的影响.针对传统水文模型参数优选采用单一目标不能充分全面挖掘水文观测资料中蕴含的水文特征信息的缺陷,本文以新安江三水源模型为例,尝试采用多目标优化算法优化率定水文模型,算例应用分析表明,通过合理的选择目标函数的种类和数目,采用多目标进化算法优化率定模型参数,可以获得相对于单目标率定模型参数更优的结果.进一步,研究工作针对模型参数优化的结果进行分析,可以明显看出模型参数优化中存在“异参同效”现象,为后续模型参数不确定性分析等相关研究工作的开展做好了铺垫.     

水文模型参数自动优选方法的比较分析

模型参数的识别是模糊研制与应用成功与否的关键。介绍了三个自动优选模型参数的方法，以新安江模型为例，应用１４个流域的资料，对罗森布郎克法、改进的单纯形法和基因算法优算法优选模型参数的效果，优化方法和收敛速度及参数初值对优选效果的影响进行了比较分析。     

基于遗传算法的新安江模型日模拟参数优选研究

在概念性水文模型的参数率定中,目前还没有一个传统优化方法能够提供保证足够高效和稳定性的算法。为了克服传统优化方法中局部收敛性的缺点,近年来利用遗传算法通过计算机准确稳定地进行概念性水文模型的参数优选的尝试得到越来越多的重视和发展。目前优选水文模型待定参数,大多是从次洪模型的方面去讨论,有关日模拟模型的遗传算法参数优选讨论的较少。本文系统分析了基于遗传算法的新安江模型日模拟参数的自动优选,同时针对遗传算法在模型参数众多的情况下时间效率低下问题,通过利用新安江模型参数分层原理与模型参数敏感性分析对优选结果影响,提出一套简化的日模型参数遗传算法优选方案。经过流域模拟检验,该优选方案可行,运行效率高,可以作为类似模型遗传算法参数率定快速、有效的方案。     

基于博弈论综合权重模糊优选模型在地下水环境风险评价中的应用

地下水环境风险评价是地下水可持续利用的前提.目前,评价方法比较单一,主要为突变理论和单因素模糊理论,它们的控制变量数有限,不适用于多目标决策,同时也没有综合考虑主客观权重,评价结果不科学.本文针对以往评价方法的不足,应用基于主客观综合权重的多层多目标模糊可变优选模型评价.模型采用熵权法计算客观权重,二元对比法确定主观权重,并利用博弈论集结模型推导计算了综合权重,并以海河流域地下水环境为例进行了风险评价.计算实例表明,该模型能较好地考虑评价指标的主客观权重,适用于多层多目标评价系统.     

基于目标函数的钻压优化模型建模方法

研究了如何把过去由地质人员离线确定优化钻压的模型变为自动送钻中在线实时的钻压优化模型。以综合钻井成本最低为准则,建立了钻井成本目标函数,讨论了影响钻井成本的因素,给出了钻压优化模型的约束条件,在以上基础上建立了钻压优化模型,并给出了钻压优化模型在线实时求解过程中有关参数的确定方法,最后总结了基于目标函数钻压优化模型建模方法优缺点和适用范围。本研究对科学自动化钻井具有参考价值。     

多目标系统模糊模型在强夯参数优选中的应用

将一种多目标系统的模糊模型用于地基强夯参数的优选，使这种定性的问题得到了定量化的处理。实例分析表明模糊优选较其它方法科学、合理。     

模糊数学方法在金刚石原生矿预测中的应用

在金刚石原生矿预测研究中应用模糊数学中的相似性优选和综合评判方法取得了好的效果。相似性优选主要利用定量数据并可直接用汉明距离求出规格化对比方阵,综合评判方法可以同时把定量和定性数据都纳入四个综合评判模型,若有必要还可以继续进行二级综合评判。在进行综合评判时必需考虑元素在岩体中的分布模型,若使用常量元素分析结果,则直接用原始数据,若为微量元素分析结果则先对原始数据先取对数。再根据样品数的多少分别选用格贴近度(N>5)和中间型隶属函数(N≤5)求值,若为其它的定性数据则采用其它相应的隶属函数求值,在此基础上再进行综合评判。笔者在实际应用时以综合结果为主,参考相似性优选结果进行选点,只利用岩石化学分析结果就获得了十分理想的效果,选出的第一点取样经鉴定即为钾镁煌斑岩。     

基于人工蜂群算法的岩体结构面多参数优势分组研究

岩体结构面优势分组是研究岩体力学性质与水力特性的基础,通常的分析方法是只根据产状进行划分。鉴于结构面其他特征对岩体力学性质的重要影响,考虑结构面倾向、倾角、迹长、张开度、表面形态5个特征参数,提出了基于人工蜂群算法的岩体结构面多参数优势分组方法。以样本总体离差平方和为目标函数,建立岩体结构面多参数优势分组的数学模型,应用人工蜂群优化算法求解,以目标函数值最小时的解作为聚类中心,并自动确定分组边界。对人工生成的结构面数据的计算结果验证了该方法的正确性,该方法的求解精度是令人满意的。最后,将该方法应用于怒江松塔水电站坝址区岩体结构面多参数优势组的划分,得到了较为合理的分组结果,进一步验证了此方法具有较高的运行效率与工程实用性。     

多目标规划在保定市水资源优化配置的应用研究

水资源短缺已成为严重制约保定市经济发展的重要因素。因此,对水资源进行优化配置研究,对水资源规划与管理有重要意义。运用多目标规划理论建立了一个考虑当地地表水、地下水及其水源多水源联合供水条件下的保定市水资源优化配置模型。模型以经济和社会的综合效益最大为目标,分析确定了模型各参数,并调用matlab优化工具箱中的函数进行编程求解,得到保定市规划水平年(2015年,2030年)的水资源优化配置方案,为该市水资源规划与管理提供依据。     

Effect of using multiple stream gauging stations on calibration of hydrologic parameters and estimation of hydrograph of ungauged neighboring basin

In this study, Hydrologic Engineering Center-Hydrologic Modeling System is used to simulate hydrologic processes in a watershed in Western Black Sea Region that frequently experiences flooding. The region is mountainous with steep hill slopes and receives high precipitation throughout the year. There are three stream gauging stations in the basin whose data are available for calibration and validation of hydrologic parameters. Simulations are performed for different scenarios to investigate the effect of using multiple stream gauging stations’ data on catchment wide calibration and validation of hydrologic parameters. Furthermore, performance of using calibrated internal stream gauging stations’ flow data in the estimation of hydrologic parameters in an assumed neighboring ungauged basin was assessed. It is found that using data of multiple stream gauging stations for calibration and validation gives satisfactory results for direct runoff hydrograph but the peak discharge predictions are not improved. The study results suggest that using data of internal stream gauging stations enables improved understanding of internal dynamics and transport in the basin and better predicted direct runoff hydrograph for the assumed neighboring ungauged basin.     

目标函数权重对PEST-HSPF水文率定的影响研究

参数率定是水文模型构建与应用的重要基础。在多目标水文校准中,不同目标函数的权重设置直接影响校准的结果,如何确定不同目标函数的权重是水文校准的关键问题。选择典型半分布式水文模型HSPF(Hydrological Simulation Program-Fortran),基于PEST多目标校准模型,以牛栏江上游流域为研究对象,研究目标函数权重对HSPF水文模拟拟合优度的影响,为HSPF模型参数自动率定提供指导与借鉴。结果表明:(1)当PEST-HSPF水文校准的单目标函数权重(日流量、月流量、超流天数)上升时,模型的纳什系数和相对偏差呈现非线性变化特征,不规则波动幅度较大;(2)3个目标函数的权重在1~10量级内模型能够获得较高的预测能力和较低的误差,模型的纳什系数平均可达到0.8以上,相对偏差在10%以内;(3)超流量天数权重设置对模型的预测能力变化影响较大,日流量天数权重对相对偏差的波动影响较大,当超流天数的相对权重在1~1 000范围内变化时,模型纳什系数在0.02~0.86之间波动剧烈,当日流量相对权重值超过30时,相对偏差变化明显。     

提孜那甫河流域融雪径流模拟及不确定性分析

为了开展寒旱山区典型流域融雪径流过程的研究，提高融雪径流模型（SRM）在山区融雪地区的水文过程模拟精度，本文选取新疆提孜那甫河流域作为典型研究区，在SRM径流计算基础上，加入合适的基流数据并进行不确定性分析。考虑4种常见的基流分割方法（数字滤波法、加里宁法、BFI法（滑动最小值法）和HYSEP（hydrograph separation program）法），基于贝叶斯理论，采用马尔科夫链蒙特卡洛（MCMC）模拟进行参数不确定性分析，对使用不同基流数据SRM的融雪径流模拟表现进行综合评价。分析结果表明，基于加里宁基流分割方法的模型（SRM_K）能够最佳地模拟研究区融雪径流过程（纳什系数NSE在识别期和验证期分别为0.866和0.721，大于其他对比模型）。MCMC模拟能够较好地识别SRM参数，获得可靠的参数后验概率分布。当实测降水资料缺乏或其代表性较差时，TRMM（tropical rainfall measuring mission）卫星数据能够描述研究区的降水过程特征。     

Improved hydrograph prediction through subsurface characterization: conditional stochastic hillslope simulations

Subsurface heterogeneity is one of the largest sources of uncertainty associated with saturated hydraulic conductivity. Recent work has demonstrated that uncertainty in hydraulic conductivity can impart significant uncertainty in runoff generation processes and surface-water flow. Here, the role of site characterization in reducing hydrograph prediction bias and uncertainty is demonstrated. A fully integrated hydrologic model is used to conduct two sets of stochastic, transient simulation experiments comprising different overland flow mechanisms: Dunne and Hortonian. Conditioning hydraulic conductivity fields using values drawn from a simulated synthetic control case are shown to reduce both mean bias and variance in an ensemble of conditional hydrograph predictions when compared with the control case. The ensemble simulations show a greater reduction in uncertainty in the hydrographs for Hortonian flow. The conditional simulations predict surface ponding and surface pressure distributions with reduced mean error and reduced root mean square error compared with unconditional simulations. Uncertainty reduction in Hortonian and Dunne flow cases demonstrates different temporal signals, with more substantial reduction achieved for Hortonian flow.     

Quantifying the effects of subsurface heterogeneity on hillslope runoff using a stochastic approach

The role of heterogeneity and uncertainty in hydraulic conductivity on hillslope runoff production was evaluated using the fully integrated hydrologic model ParFlow. Simulations were generated using idealized high-resolution hillslopes configured both with a deep water table and a water table equal to the outlet to isolate surface and subsurface flow, respectively. Heterogeneous, correlated random fields were used to create spatial variability in the hydraulic conductivity. Ensembles, generated by multiple realizations of hydraulic conductivity, were used to evaluate how this uncertainty propagates to runoff. Ensemble averages were used to determine the effective runoff for a given hillslope as a function of rainfall rate and degree of subsurface heterogeneity. Cases where the water table is initialized at the outlet show runoff behavior with little sensitivity to variance in hydraulic conductivity. A technique is presented that explicitly interrogates individual realizations at every simulation timestep to partition overland and subsurface flow contributions. This hydrograph separation technique shows that the degree of heterogeneity can play a role in determining proportions of surface and subsurface flow, even when effective hillslope outflow is seen. This method is also used to evaluate current hydrograph separation techniques and demonstrates that recursive filters can accurately proportion overland and base-flow for certain cases.     

新安江产流模型与改进的BP汇流模型耦合应用

为提高新安江模型的汇流计算精度并减少经验因素对参数率定的影响,将新安江产流模型与改进的BP汇流模型相耦合,建立XBK(XAJ-BP-KNN)模型。该模型以前期模拟流量和新安江产流模型计算的产流量作为BP网络的输入,出口断面流量作为网络输出,拟合汇流的非线性关系,代替新安江模型的分水源、线性水库及河道马斯京根法的汇流计算;采用相似原理和K-最近邻算法,基于历史样本的模拟误差及相应影响要素对网络输出进行误差修正,实现了无前期实测流量的连续模拟;模型使用SCE-UA算法与遗传早停止LM算法相结合的全局优化方法进行参数优选。在呈村流域的验证表明XBK模型的模拟精度高于新安江模型,全局优化方法能找到最优参数,降低了模型的使用难度。     

基于泉流量过程线的管道流与裂隙流识别

岩溶含水系统存在不规则的裂隙网络和管道结构,由于贮水空间的差异,导致水流运移的多变性和水动力参数的复杂难测。基于泉流量过程线的分析技术能够很好地识别岩溶系统中的管道流和裂隙流。文章将泉流量分为3个组成部分:前期蓄水量、快速径流和慢速径流,其中快速径流和慢速径流视为两线性并联水库,建立了前期降雨和瞬时单位线之间的识别函数,识别降雨在岩溶不同空隙类型中的水量分配系数,并将该函数模型应用到后寨地下河岩溶小流域。模拟结果表明,上游到下游前期蓄水量呈增加趋势,而水量进入管道中的分配系数分别为0.84,0.6以及0.48,呈现递减趋势,说明岩溶裂隙结构越往下游越发育。模型计算出的泉流量曲线与实测值相比,NSE达到90%以上,模拟效果理想。     

Submarine and intertidal groundwater discharge through a complex multi-level karst conduit aquifer

The quantification of submarine and intertidal groundwater discharge (SiGD) or purely submarine groundwater discharge (SGD) from coastal karst aquifers presents a major challenge, as neither is directly measurable. In addition, the expected heterogeneity and intrinsic structure of such karst aquifers must be considered when quantifying SGD or SiGD. This study applies a set of methods for the coastal karst aquifer of Bell Harbour in western Ireland, using long-term onshore and offshore time series from a high-resolution monitoring network, to link catchment groundwater flow dynamics to groundwater discharge as SiGD. The SiGD is estimated using the “pollution flushing model”, i.e. a mass-balance approach, while catchment dynamics are quantified using borehole hydrograph analysis, single-borehole dilution tests, a water balance calculation, and cross-correlation analysis. The results of these analyses are then synthesised, describing a multi-level conduit-dominated coastal aquifer with a highly fluctuating overflow regime draining as SiGD, which is in part highly correlated with the overall piezometric level in the aquifer. This concept was simulated using a hydraulic pipe network model built in InfoWorks ICM [Integrated Catchment Modeling]^® version 7.0 software (Innovyze). The model is capable of representing the overall highly variable discharge dynamics, predicting SiGD from the catchment to range from almost 0 to 4.3 m³/s. The study emphasises the need for long-term monitoring as the basis for any discharge studies of coastal karst aquifers. It further highlights the fact that multiple discharge locations may drain the aquifer, and therefore must be taken into consideration in the assessment of coastal karst aquifers.     

Continuous permeability predictions in heterogeneous reservoirs using Vshale and microstructure calibrated free-fluid models (a combined study of a Niger Delta field and the tight gas sandstone of the southern North Sea)

The well-established free-fluid model from NMR technique provides continuous permeability values that closely match with core permeabilities better than most theoretical models especially when it is core calibrated for field specific use. However, only few wells have NMR logs in a field while marginal fields may not have any due to economic reasons. This study explored means of achieving one of the overriding objectives of most marginal field operators, which is to reduce the overall cost of production to the attainable minimum. The free-fluid model was modified into two simple and cost-effective models in order to optimize its applicability to predict permeability in the absence of NMR data. The two new models, which were developed for the single and double porosity systems analyzed in this study, consist of calibration parameters that can be empirically determined to account for variation in reservoir quality based on the rock type profile per field. A non-matrix parameter, α, was introduced into the model derived for tight gas sandstone being regarded as a double-porosity formation. This inclusion represents the permeability contribution of natural fractures or any crack-like pores to the different flow units. By using the alternative version to the known free-fluid model, continuous permeability curves that match experimental results were predicted without NMR logs.     

Effect of watershed partitioning on hydrologic parameters and estimation of hydrograph of an ungauged basin: a case study in Gokirmak and Kocanaz,Turkey

The main goal of this study is to investigate the effect of the size of the subbasins of a watershed on the hydrologic parameters and their spatial variability in an estimation of the hydrologic parameters and hydrograph of a neighbouring ungauged basin. In this paper, Hydrologic Engineering Center-Hydrologic Modelling System (HEC-HMS), a semi-distributed hydrologic model, is used to calibrate and cross-validate two flood events occurred in 1998 and then validate four other flood events occurred in 1991, 1994, 2002, and 2009 in Gokirmak Basin in Western Black Sea Region, Turkey. The basin is divided into seven different subbasins to investigate the effect of watershed partitioning on calibrated hydrologic parameters of each subbasin using the peak-weighted root mean square error method as an objective function and the hydrograph at the outlet of the whole basin. It is found out that as the geometric magnitudes of the subbasins changed, the calibrated values of the hydrologic parameters of those subbasins changed as well. Then, a neighbouring basin, Kocanaz, is considered as an assumed neighbouring ungauged basin to investigate the effect of watershed partitioning of a gauged basin on the estimation of hydrograph of a neighbouring ungauged basin. Hydrologic parameters and direct runoff hydrograph of assumed ungauged neighbouring basin are estimated from the hydrologic parameters of the HEC-HMS calibration results of Gokirmak. Statistical indicators of the simulation results for each basin partitioning were graded with respect to the boundary values of the simulation outputs to find the best alternative. The grading results show that the simulation results with a single basin gave better representation among all other partitioning except two flood events.