A Comprehensive Python Toolkit for Accessing High‐Throughput Computing to Support Large Hydrologic Modeling Tasks

The National Flood Interoperability Experiment (NFIE) was an undertaking that initiated a transformation in national hydrologic forecasting by providing streamflow forecasts at high spatial resolution over the whole country. This type of large‐scale, high‐resolution hydrologic modeling requires flexible and scalable tools to handle the resulting computational loads. While high‐throughput computing (HTC) and cloud computing provide an ideal resource for large‐scale modeling because they are cost‐effective and highly scalable, nevertheless, using these tools requires specialized training that is not always common for hydrologists and engineers. In an effort to facilitate the use of HTC resources the National Science Foundation (NSF) funded project, CI‐WATER, has developed a set of Python tools that can automate the tasks of provisioning and configuring an HTC environment in the cloud, and creating and submitting jobs to that environment. These tools are packaged into two Python libraries: CondorPy and TethysCluster. Together these libraries provide a comprehensive toolkit for accessing HTC to support hydrologic modeling. Two use cases are described to demonstrate the use of the toolkit, including a web app that was used to support the NFIE national‐scale modeling.     

基于Blender的矿山虚拟现实快速构建与实现

为了实现矿山三维实体模型和虚拟现实的快速搭建与实现,分析了当前数字化矿山建模可移植性差、扩展性不灵活、开发难度大的缺陷 ,提出了一种矿山三维模型及虚拟现实体系的自动构建方法:基于Blender开源软件提供的建模平台,利用Python语言编写矿山实体的表达方法 与生成算法;利用矿山基础信息,实现了对矿山井巷、矿体等三维实体模型的自动、快速构建,并以矿山井巷工程、矿体以及地表模型等实体 模型的构建为例进行了演示;利用Blender脚本实现对三维实体模型的快速渲染,初步建立了矿山虚拟现实环境与仿真系统,并利用驱动内核实 现虚拟情景的演示。结果表明,该方法具有开发快捷、要求低等特点,为数字化矿山建设提供了一种新的思路。     

基于Adam优化深度神经网络快速确定瓦斯抽采半径*

为解决煤矿瓦斯有效抽采半径难以快速准确确定的问题,采用基于Adam算法优化DNN（深度神经网络）方法来预测瓦斯抽采半径。查阅文献共收集已得到验证的970组数据集,每组数据选取煤层瓦斯初始渗透率、钻孔直径、抽采时间、地应力、煤层初始瓦斯压力作为预测模型的5个特征量,有效抽采半径作为目标输出值。接着预测模型进行不断学习和训练,最终训练得到1个最优的瓦斯有效抽采半径预测模型。利用训练好的最优预测模型结合Python语言开发出计算有效抽采半径的软件,并使用该软件在四季春煤矿和鹤煤六矿进行有效抽采半径预测的工程实例研究,验证该软件预测抽采半径的实用性和准确性。研究结果表明:通过使用开发的软件,可快速且较准确地计算出矿井瓦斯有效抽采半径,可为暂不具备现场测试条件的矿井抽采设计提供一定的参考依据。     

基于Python的ABAQUS二次开发在复合材料层合板冲击建模中的应用

目的为了简化复合材料层合板冲击建模过程,节省模型处理时间,利用python语言进行ABAQUS二次开发。方法通过编写主控文本程序Damagemodel.py,图形界面编辑程序DamagemodelDB.py和衔接程序Damagemodel_plugin.py,实现系统建模界面定制,复合材料层合板和弹头的快速建模参数化,以及分析步和接触属性定义自动化。结果在快速建模完成的基础上,对冲击过程进行模拟。冲击过程中,层合板冲击面产生凹坑,背部凸起。仿真深度与实测深度误差为0.2 mm,损伤面积仿真结果与试验结果误差为4.2%。结论模拟结果与试验结果较为接近,表明快速建模可以为冲击模拟提供有效的支持。     

PyLUR: Efficient software for land use regression modeling the spatial distribution of air pollutants using GDAL/OGR library in Python

• PyLUR comprises four modules for developing and applying a LUR model. • It considers both conventional and novel potential predictor variables. • GDAL/OGR libraries are used to do spatial analysis in the modeling and prediction. • Developed on Python platform, PyLUR is rather efficient in data processing. Land use regression (LUR) models have been widely used in air pollution modeling. This regression-based approach estimates the ambient pollutant concentrations at un-sampled points of interest by considering the relationship between ambient concentrations and several predictor variables selected from the surrounding environment. Although conceptually quite simple, its successful implementation requires detailed knowledge of the area, expertise in GIS, statistics, and programming skills, which makes this modeling approach relatively inaccessible to novice users. In this contribution, we present a LUR modeling and pollution-mapping software named PyLUR. It uses GDAL/OGR libraries based on the Python platform and can build a LUR model and generate pollutant concentration maps efficiently. This self-developed software comprises four modules: a potential predictor variable generation module, a regression modeling module, a model validation module, and a prediction and mapping module. The performance of the newly developed PyLUR is compared to an existing LUR modeling software called RLUR (with similar functions implemented on R language platform) in terms of model accuracy, processing efficiency and software stability. The results show that PyLUR out-performs RLUR for modeling in the Bradford and Auckland case studies examined. Furthermore, PyLUR is much more efficient in data processing and it has a capability to handle detailed GIS input data.     

基于机器学习的浮标实时自动化赤潮预警研究

实现赤潮预警对于减轻海洋环境灾害、避免海洋产业特别是海洋渔业重大经济损失具有重要意义。针对当前水文监测数据海量却难以实现实时自动化监测与预警,特别是难以利用传统监测手段实现对危害更大的赤潮的精准实时预测这一显著问题,提出利用浮标数据作为依据,借助机器学习在大数据分析和智能决策方面的优势,建立一种新颖的双重递进式赤潮预警机制的方法。首先,通过相关算法分析历史数据,以确认赤潮初步预警阈值;其次,对叶绿素a、pH、溶解氧等重要监测指标的当前和阶段性变化进行初步分析,判断是否达到预警触发条件;然后,进一步联合分类、回归、聚类、神经网络等机器学习相关方法,对数据进行深度挖掘;最后,通过这种递进式的机制对短期内是否会发生赤潮作出判断,以实现赤潮自动化预警预报。在此基础上,利用宁波梅山湾实际监测数据,证实了该方法在赤潮实时自动化预警中的有效性。     

基于Keras长短时记忆网络的矿井瓦斯浓度预测研究

瓦斯浓度监测是煤矿瓦斯灾害事故预警的重要的手段,其浓度变化预测对于提升矿山安全生产具有重要意义。针对矿井瓦斯浓度预测问题,建立了一种基于Keras长短时记忆网络的矿井瓦斯浓度预测模型。该模型首先对矿井瓦斯浓度时间序列进行标准化处理,并将处理后的时间序列划分为训练集与测试集;然后通过调用测试集数据进行模型训练,利用提出的基于LSTM网络建立的矿井瓦斯浓度多步预测模型,实现了对矿井瓦斯浓度发展趋势的预测,并利用损失函数计算预测误差大小,评估模型的预测精度;最后以贵州某煤矿掘进工作面为工程背景,利用基于LSTM网络建立的矿井瓦斯浓度多步预测模型,开展了矿井瓦斯浓度预测研究,并通过与ARMA模型、ARIMA模型的预测结果进行对比,验证该模型的预测效果。结果表明:该模型预测结果的均方根误差RMSE值最小仅为2%,且预测步长约为ARMA模型、ARIMA模型的5倍,说明该模型的预测效果好,可为煤矿井下合理规避瓦斯灾害事故提供科学依据。