基于多源数据融合的煤矿工作面瓦斯浓度预测*

为了解决瓦斯浓度预测使用的单一数据在预测中影响还不够深入的问题,提出基于LSTM神经网络的多源数据融合瓦斯浓度预测模型。模型将上隅角瓦斯浓度、采煤机速度、工作面吨煤瓦斯涌出量等不同数据融合作为输入层参数,使用Adam优化算法更新LSTM网络层参数,利用Attention机制突出关键影响瓦斯浓度的因素,开展多源数据融合的瓦斯浓度预测,结合某矿1008工作面的实际数据,分析不同数据在瓦斯浓度预测中的作用。研究结果表明:单变量下的Attention-aLSTM预测效果相比LSTM提升14.2%;多源数据融合下的Attention-aLSTM相比自身提升了5%。     

土地资源的多级网格数据结构建立与应用研究

传统的基于行政区的土地统计数据不能完全表现区域内部土地利用的空间分异特征,以武汉市为实验区,对基于网格的统计信息算法STING(Statistical Information Grid-based method)进行扩展,以景观多样性指数为定量化指标对实验区进行四叉树划分生成不均匀多级网格,建立一种拟合了行政区划界线的不均匀的多级网格结构来存储、管理和分析土地数据。并以此多级网格数据结构为平台计算和生成实验区人口密度空间分异渲染图,初步抽取了人口分布与土地利用之间的关系。实验表明,基于多级网格的统计方法能更好地表达土地利用及其相关数据的空间分异性,利于对土地资源数据的进一步挖掘以抽取所需知识。     

Impact of bias in predicted height on tree volume estimation: A case-study of intrinsic nonlinearity

Bias originating from intrinsic nonlinearity in nonlinear models is caused by excess curvature in the solution locus of parameter estimates derived from least squares procedures. Bias due to intrinsic nonlinearity varies according to sample size as well as model specification. This paper analyses consequences of fractionising data into smaller sub-samples. Based on measurements of stem diameter and total tree height from the first Danish national forest inventory, it is demonstrated how data splitting at random may cause the intrinsic nonlinear curvature to exceed the critical F-value. Application of a Taylor-series expansion shows that, for all practical purposes, the bias in predictions of individual tree volume (based on stem diameter and tree height) is negligible. To minimize residual variance, intrinsic curvature and, in turn, prediction bias, it is recommended that data be stratified according to site conditions, stand characteristics or other relevant criteria. Finally, the preferred model should exhibit close-to-linear behaviour.     

Modelling the forest carbon budget of a Mediterranean region through the integration of ground and satellite data

This paper introduces an innovative modelling strategy aimed at simulating the main terms of net forest carbon budget (net primary production, NPP and net ecosystem exchange, NEE) in Tuscany (Central Italy). The strategy is based on the preliminary calibration and application of parametric and bio-geochemical models (C-Fix and BIOME-BGC, respectively), which simulate the behaviour of forest ecosystems close to equilibrium condition (climax). Next, the ratio of actual over-potential tree volume is computed as an indicator of ecosystem distance from climax and is combined with the model outputs to estimate the NPP and NEE of real forests. The per-pixel application of the new modelling strategy was made possible by the collection of several data layers (maps of forest type and volume, daily meteorological data and monthly normalized difference vegetation index (NDVI) images for the years 1999–2003) which served to characterize the eco-climatic and forest features of the region. The obtained estimates of forest NPP and NEE were evaluated against ground measurements of accumulated woody biomass and net carbon exchange. The results of these experiments testify the good potential of the proposed strategy and indicate some problem areas which should be the subject of future research.     

Bootstrap methods for simultaneous benchmark analysis with quantal response data

A primary objective in quantitative risk assessment is the characterization of risk which is defined to be the likelihood of an adverse effect caused by an environmental toxin or chemcial agent. In modern risk-benchmark analysis, attention centers on the “benchmark dose” at which a fixed benchmark level of risk is achieved, with a lower confidence limits on this dose being of primary interest. In practice, a range of benchmark risks may be under study, so that the individual lower confidence limits on benchmark dose must be corrected for simultaneity in order to maintain a specified overall level of confidence. For the case of quantal data, simultaneous methods have been constructed that appeal to the large sample normality of parameter estimates. The suitability of these methods for use with small sample sizes will be considered. A new bootstrap technique is proposed as an alternative to the large sample methodology. This technique is evaluated via a simulation study and examples from environmental toxicology.

水环境监测结果的审核重点与审核方法

环境水质监测分析过程中水质指标多,监测数据多,数据的审核工作比较复杂。数据审核是水质分析工作质量保证的一个重要环节,是整个质量保证体系中有效的控制手段。主要从理论上分析环境监测中水质指标之间的关系,为水质监测人员监测和数据审核提供参考,从而提高数据质量,为环境水质管理服务。     

生命周期清单分析的数据质量评价

生命周期清单分析(LCI)数据质量的分析方法可概括为两类:采用诸如地理代表性、数据年代或数据获取方式等一系列指标来表示;根据不确定性来综合反映LCI质量.笔者在分析了这两类方法各自所存在缺陷的基础上,提出了将这两者相结合的评价方法:采用5个独立的反映数据质量的指标,根据系统各单元各数据属性对各指标从1～5进行打分,形成数据的质量指标向量元素.根据数据质量向量元素的算术平均在总指标范围中所占的百分数将质量指标向量转化为对应的综合数据质量指标(DQI),继而根据DQI可得出每个数据的随机分布,以便进行清单结果的不确定性随机模拟.最后将方法应用于钢铁生产生命周期清单数据中.      

Stockhome: A Spreadsheet Model of Urban Heavy Metal Metabolism

Computer models for analysis,visualising and decision support in environmentalresearch have become increasingly popular. TheStockhome project, where the urban metabolism ofheavy metals in Stockholm was studied, resultedin a database with historical data of the use ofgoods containing cadmium (Cd), chromium (Cr),copper (Cu), lead (Pb), mercury (Hg), nickel (Ni)and zinc (Zn). A spreadsheet model was developedto study flows and stocks of the metalconsumption process and emissions. The modelindicates uncertainties of the data, societalaspects such as field of use and rights ofdisposition of the goods. By considering goods asthe drivers of the emissions, the model would bewell suited for policy support.     

ENHANCEMENTS OF NONPOINT SOURCE MONITORING OF VOLATILE ORGANIC COMPOUNDS IN GROUND WATER1

ABSTRACT: The U.S. Geological Survey (USGS) has compiled a national retrospective data set of analyses of volatile organic compounds (VOCs) in ground water of the United States. The data are from Federal, State, and local nonpoint‐source monitoring programs, collected between 1985–95. This data set is being used to augment data collected by the USGS National Water‐Quality Assessment (NAWQA) Program to ascertain the occurrence of VOCs in ground water nationwide. Eleven attributes of the retrospective data set were evaluated to determine the suitability of the data to augment NAWQA data in answering occurrence questions of varying complexity. These 11 attributes are the VOC analyte list and the associated reporting levels for each VOC, well type, well‐casing material, type of openings in the interval (screened interval or open hole), well depth, depth to the top and bottom of the open interval(s), depth to water level in the well, aquifer type (confined or unconfined), and aquifer lithology. VOCs frequently analyzed included solvents, industrial reagents, and refrigerants, but other VOCs of current interest were not frequently analyzed. About 70 percent of the sampled wells have the type of well documented in the data set, and about 74 percent have well depth documented. However, the data set generally lacks documentation of other characteristics, such as well‐casing material, information about the screened or open interval(s), depth to water level in the well, and aquifer type and lithology. For example, only about 20 percent of the wells include information on depth to water level in the well and only about 14 percent of the wells include information about aquifer type. The three most important enhancements to VOC data collected in nonpoint‐source monitoring programs for use in a national assessment of VOC occurrence in ground water would be an expanded VOC analyte list, recording the reporting level for each analyte for every analysis, and recording key ancillary information about each well. These enhancements would greatly increase the usefulness of VOC data in addressing complex occurrence questions, such as those that seek to explain the reasons for VOC occurrence and nonoccurrence in ground water of the United States.     

基于随机森林的阻变型通风网络故障诊断方法*

为了在矿井通风网络发生阻变型故障时,能够快速准确判断出故障位置和故障量,提出1种基于随机森林的通风网络故障位置和故障量诊断方法。利用矿井通风仿真系统IMVS将唐安矿模拟故障生成空间数据集并进行数据预处理,构建基于随机森林的故障诊断模型,并利用该诊断模型对唐安矿矿井通风网络模拟故障位置和故障量进行判断和预测。引用多种方法对模型进行度量,通过唐安矿模拟实验验证基于随机森林的故障诊断模型的有效性。将随机森林和决策树的故障诊断准确率进行对比,研究结果表明:随机森林较决策树故障准确率有进一步的提高,并发现故障地点失误诊断多是相邻巷道,在一定程度上工作人员对故障地点的判断并不受其影响。