基于BERT-BILSTM-CRF模型的电力行业事故文本智能分析*

为解决电力行业事故报告文本较长、语义复杂,难以进行有效文本识别问题,提出1种以BERT作为底层的预训练模型,并设计1种双重注意力机制编码器,结合BILSTM-CRF深度挖掘事故文本语义特征,从而实现文本智能分析。首先构建电力词典,通过对BERT预训练,进行BIO标注,然后引入BILSTM-CRF模型实现对文本标签智能分类,最后将该模型与现行其他4种深度学习模型进行对比。研究结果表明:该模型智能识别精确率、召回率及F1值(查准率)均达到约97%,较其他4种模型中效果最好的模型分别提高0.02,0.03,0.02。研究结果可为电力行业事故报告文本分析提供1种新思路。     

Resident participation in post-Lushan earthquake housing reconstruction: a multi-stage field research method-based inquiry

In April 2013, a 7.0?Mw earthquake struck in Lushan County, Sichuan Province, China, causing serious housing damage. Housing reconstruction is a crucial part of disaster recovery, and resident participation has been recognized as a key part of reconstruction success. This paper examined housing reconstruction resident participation in the post-Lushan earthquake period as a multi-stage problem based on multi-stakeholder collaborative perspectives, which covered emergency shelter, temporary housing and permanent housing stages. Multi-stage field research was conducted in the affected areas, and the residents were found to be fully involved in the reconstruction with a ‘high – medium – high' distribution across the three stages. Resident participation experiences were summarized in terms of previous earthquake education, NGO participation and community-based organization functions, and the reasons discussed, such as civic awareness and civic rights. Problems regarding improper reconstruction and cultural conflicts were also examined and valuable suggestions given.     

Role of goethite dissolution in the oxidation of 2-chlorophenol with hydrogen peroxide.

It is well known that the dissolution of goethite plays an important role in catalyzing the oxidation of organic chemicals. Therefore, this study investigates how surface dissolution of goethite affects 2-chlorophenol oxidation in the goethite/H2O2 process. Experimental results indicate that ligand and reductant can enhance the dissolution rate of goethite, which is surface-controlled. Our results further indicate 2-chlorophenol degradation depends on goethite concentration. In addition, the oxidation rate of 2-CP is correlated with reductive dissolution rate at various dosages of goethite. Moreover, the oxidation mechanism of 2-CP is also a surface-controlled reaction. A mechanism proposed herein indicates that, in addition to the contaminant, its intermediate species affect the oxidation rate as well.     

Geostatistical analysis and risk assessment on soil total nitrogen and total soil phosphorus in the Dongting Lake plain area, China

Nonpoint-source pollution and water body eutrophication have become increasing concerns for scientists and policymakers. Nitrogen and phosphorus affect environmental pollution, especially lake eutrophication. To assess the environmental risk of soil total nitrogen (TN) and total phosphorus (TP) pollution, a typical ecological unit of Dongting Lake plain was selected as the experimental site. To verify the stationary of the data, a moving windows technique was adopted. Our results showed that Box-Cox transformation achieved normality in the data set and dampened the effect of outliers. The best theoretical model for semivariogram of TN and TP was a spherical model. The ordinary kriging estimates of TN and TP concentrations were mapped. The integrative comparisons of semivariogram parameters with different trends to the kriging prediction errors of TN and TP indicated that the two-order trend is preferable. Kriging SDs provided valuable information that will increase the accuracy of TN and TP mapping. The probability kriging method is useful to assess the risk of N and P pollution by providing the conditional probability of N and P concentrations exceeding the threshold concentrations of 3.2 and 1.05 g/kg, respectively. The probability distribution of TN and TP at different levels will be helpful to conduct risk assessment, optimize fertilization, and control the pollution of N and P.     

Vegetation and carbon sequestration and their relation to water resources in an inland river basin of Northwest China

In the Heihe River Basin in the arid inland area of northwest China, the distribution of water resources in vegetation landscape zones controls the ecosystems. The carbon sequestration capacity of vegetation is analyzed in relation to water resources and vegetation growing conditions. During the last 20 years, the vegetation ecosystems have degenerated in the Heihe River Basin. Simulation using the C-FIX model indicates that, at present, the total amount of NPP of vegetation accounts for about 18.16 TgC, and the average value is 106 gC/m(2)/yr over the whole basin. NPP has generally the highest value in the upperstream mountain area, middlestream artificial oases area, downstream river bank area, alluvial fan and the terminal lake depression where vegetation grows relatively well. The lowest value is found in the vast downstream desert and Gobi area. Protection of vegetation ecosystems and enhancement of carbon sequestration require such inland river basins as the Heihe River Basin to be brought under management in a comprehensive way, taking water as a key, to carry out a rational and efficient allocation and utilization of water resources.     

Water Hyacinth in China: A Sustainability Science-Based Management Framework

The invasion of water hyacinth (Eichhornia crassipes) has resulted in enormous ecological and economic consequences worldwide. Although the spread of this weed in Africa, Australia, and North America has been well documented, its invasion in China is yet to be fully documented. Here we report that since its introduction about seven decades ago, water hyacinth has infested many water bodies across almost half of China’s territory, causing a decline of native biodiversity, alteration of ecosystem services, deterioration of aquatic environments, and spread of diseases affecting human health. Water hyacinth infestations have also led to enormous economic losses in China by impeding water flows, paralyzing navigation, and damaging irrigation and hydroelectricity facilities. To effectively control the rampage of water hyacinth in China, we propose a sustainability science-based management framework that explicitly incorporates principles from landscape ecology and Integrated Pest Management. This framework emphasizes multiple-scale long-term monitoring and research, integration among different control techniques, combination of control with utilization, and landscape-level adaptive management. Sustainability science represents a new, transdisciplinary paradigm that integrates scientific research, technological innovation, and socioeconomic development of particular regions. Our proposed management framework is aimed to broaden the currently dominant biological control-centered view in China and to illustrate how sustainability science can be used to guide the research and management of water hyacinth.     

基于“人工智能”驱动的农村生活污水系统治理专栏序言

伴随流域水污染控制与治理工作的深入,农村分散生活污水治理逐渐成为制约流域水环境改善的关键.同时,农村生活污水治理是人居环境提升、美丽乡村建设以及乡村振兴战略实施的集中体现,事关广大农村居民的切身利益.本专栏是国家水专项“分散污水治理课题”团队近年来科研成果的总结,从污水治理的全生命周期角度出发,尝试通过人工智能赋能治理模式决策、“菜单式”工艺组合、设施运维与监管、长效管理等4个方面对研究和实践进行了梳理总结,构建基于人工智能驱动的农村生活污水治理决策支持系统,以期为北运河流域以及其他区域农村生活污水治理提供科技支撑.     

电活性生物膜阳极构建及其对2，4，6-三氯酚同步降解和产电

为实现氯酚（CPs）的高效降解和资源化利用,探究微生物燃料电池（MFC）体系优势功能菌,揭示生物降解路径.接种、驯化长春市南郊污水处理厂的厌氧活性污泥,获得生物膜阳极以构筑MFC-2,4,6-TCP体系,基于扫描电子电镜（SEM）、16S rRNA分析测序方法,考察生物膜阳极微生物的附着情况和优势菌种,基于电化学阻抗（EIS）、循环伏安（CV）和线性扫描伏安（LSV）等电化学分析手段,表征生物阳极的电化学性能和氧化还原能力.结果表明,生物膜阳极微生物种类丰富,其中Geobacter和Acinetobacter分别为MFC-2,4,6-TCP体系产电和降解驯化期的优势功能菌,体系最大输出电压可达0.55 V,最大功率密度为428.65 mW·m-2,对2,4,6-TCP的降解和矿化率可达97.5%和85.4%.随着MFC循环次数的增加,微生物代谢途径多样化,产电菌逐渐演替为协同菌,且优势功能菌对2,4,6-TCP降解的中间产物（环己醇）,其毒性远低于氯酚或苯酚,更利于被微生物利用.该结果可为氯酚废水的实际处理提供新策略和技术参考.     

2021年3月14—16日中国北方地区沙尘暴天气过程诊断及沙尘污染输送分析

2021年3月14—16日发生在北方地区的沙尘暴天气过程被认为是近10年来我国出现的最强过程.本文在对此次沙尘暴过程天气学特征分析的基础上,利用 HYSPLIT 模式和 GDAS 资料,运用潜在源贡献因子分析法（PSCF）和浓度权重轨迹分析法（CWT）,探讨了气候变化背景下沙尘输送源与天气系统的配置关系及本次过程中沙尘污染物PM₁₀质量浓度的潜在源区分布及贡献.结果表明,此次过程由强烈发展的蒙古气旋及冷锋过境,高空槽后冷空气持续补充引起,中高层强斜压性使地面蒙古气旋强烈发展,大风卷扬起的沙尘随上升气流输送到高空,并在偏北大风引导下,影响了我国大范围地区.内蒙古东北部至河套地区的强涡度梯度带、500、700 hPa较高的强冷平流中心与下层的温度平流差异以及交替的上升下沉运动为本次北方地区大范围沙尘暴过程提供了动力、热力及不稳定度条件.本次沙尘天气过程中,影响呼和浩特、北京的沙尘传输通道主要为北偏东路,影响银川的沙尘传输通道为西北路和北路,过程受多沙源传输通道影响.萨彦岭、蒙古国南部戈壁沙漠为本次沙尘天气PM₁₀的主要潜在源区,传输过程中混合内蒙古沙源地沙尘.总体来说,蒙古国南部戈壁沙漠对本次过程PM₁₀质量浓度的贡献最大.     

基于无人机影像的城市植被精细分类

针对传统植被资源调查方法工作量大、成本高、效率低的问题,利用高分辨率无人机遥感影像,联合地物光谱-纹理-空间信息,构建了一种适用于描述城市不同植被种类的多维特征空间,在此基础上对三种应用广泛的分类算法(基于像素的、面向对象的支持向量机及深度学习Mobile-Unet语义分割模型)开展了对比分析研究.结果表明:本文提出的联合地物光谱-纹理-空间信息的特征空间构建方法能够有效地描述城市不同类型植被的特征差异,提升影像分割、植被分类的精度;在分类精度上,基于像素和面向对象的支持向量机分类结果的总体精度均超过90%,深度学习方法的总体分类精度为84%;在算法效率上,传统机器学习方法也优于深度学习方法.因此,得出结论针对城市小区域、小样本的植被精细分类,传统机器学习分类方法比深度学习方法效果更好.