基于神经网络分类器的城市污水处理厂水力冲击预警系统

提出了一套基于神经网络分类器的城市污水处理厂水力负荷冲击预警系统,以期对进水水量骤增现象进行提前1天的预报,使污水处理厂可根据预报结果提前采取水力冲击防护措施,从而保证各单元的平稳运行.根据进水水量的涨幅将某污水处理厂12年日进水水量监测数据分为"常规"和"冲击"两类,重点对"冲击"数据进行提前1天的预测,并采用冲击漏报率、冲击误报率和报准率对模型的预测精度进行评价;同时,基于同样的建模方法和不同的训练、验证样本建立了N(1)、N(2)和N(3)3个平行模型,以对模型的鲁棒性和建模方法的可重复性进行考察.结果显示,3个模型对2010年、2011年和2012年3年测试样本的预测效果良好,冲击漏报率和报准率两项指标数值均较为稳定,分别在0~0.167和0.981~0.995之间浮动,冲击误报率虽然在数值上的浮动较大,最低为0.143,最高为0.500,平均为0.310,但仍在工程上的可承受范围内.该结果表明,本研究基于神经网络分类器所建立的3个神经网络模型预测精度高、鲁棒性好,显示出良好的性能,有望为污水处理厂水力冲击防护工作提供有力参考.     

我国生物质气化发电行业综合风险评价

针对我国生物质气化发电行业的现状,总结诸多风险的影响因素,建立了我国生物质气化发电行业综合风险评价体系。该评价体系主要包括5个一级风险影响因素:市场风险、技术风险、安全风险、管理风险、环境风险,每个一级影响因素包括4个二级风险影响因素。由层次分析法获得各个一级影响因素和二级影响因素的权重值,根据所构建的评价集和评分标准,结合模糊综合评价法对我国生物质气化发电行业现状的综合风险等级做定量定性分析。综合风险评价研究结果表明,生物质气化发电行业处于"较高"的风险等级,与实际情况相符。综合风险评价体系的构建合理,所建立的评价方法以及相应对策分析可为决策者提供参考依据。     

Piriformospora indica confers cadmium tolerance in Nicotiana tabacum

Piriformospora indica, a root-colonizing endophytic fungus of Sebacinales, promotes plant growth and confers resistance against biotic and abiotic stresses. In order to confirm the influence of P. indica on growth, proline, malondialdehyde (MDA), chlorophyll, and cadmium (Cd) amounts in Nicotiana tabacum under Cd stress, hydroponics, pot and field trials were conducted. The results showed that P. indica can store Cd in plant roots and reduce leaf Cd content, reduce the concentration of MDA, and increase the proline and chlorophyll content and the activities of catalase, peroxidase, and superoxide dismutase under hydroponic Cd stress. RT-PCR analysis showed that the relative expression level of genes Gsh2, TaPCS1, oas1, GPX, and Hsp70 in colonized plants was 4.3, 1.4, 2.9, 1.7, and 6.9 fold higher than in un-colonized plants respectively. Cd exposure significantly reduced un-colonized plants'' agronomic traits compared to P. indica-colonized ones. Our results suggested that P. indica can sequester Cd in roots, so that much less cadmium was transported to leaves, and the increased concentrations of antioxidant enzymes, pigments and proline contents, as well as the higher expression of stress-related phytochelatin biosynthesis genes in P. indica-inoculated plants, may also serve to protect N. tabacum plants against oxidative damage, enhancing Cd tolerance.     

太阳能驱动的一体化设备在农村污水处理中的应用

以太湖流域环境敏感区内某村为例,介绍了太阳能驱动的污水处理一体化设备对农村生活污水处理的应用。连续6个月的运行及监测数据表明,污水处理实际消耗电量与光伏发电量比值为54.3%,能够有效地保证污水处理所需动力。污水处理部分采用A/O-BF一体化技术,生活污水经处理以后,出水水质达到GB 18918-2002《城镇污水处理厂污染物排放标准》中一级A标准。太阳能驱动的污水处理一体化设备运行稳定、故障率低、无需投加药剂且无运行费用,适宜在太湖流域的农村地区推广应用。     

基于分形维数的斜坡非饱和带土体大孔隙分布研究

对取自于云南省昭通市头寨滑坡滑源区附近的两个大尺度(25cm×25cm×50cm)土柱样品进行CT扫描,选取典型剖面切片,通过CT值范围将大孔隙分离从而研究土体中的大孔隙分布;计算大孔隙分布分维数(盒维数),研究其大孔隙的不规则特征,结果表明:土体中大孔隙的盒维数随深度增加呈单调递减规律,且与大孔隙度具有显著正相关关系;但大孔隙度只能表征在某一深度上大孔隙的分布多少,而大孔隙的不规则特征则需盒维数表征;最后得到大孔隙的累积大孔隙度与大孔隙孔径(在盒维数为常数时)符合幂函数规律,研究结果有利于利用分形理论建立大孔隙模型,从而研究水分在大孔隙中运动过程。     

对核磁共振成像仪建设项目环境影响评价的研究

核磁共振作为一种分析手段广泛应用于物理、化学、生物、医疗等领域.本文从核磁共振原理出发,分析核磁共振成像仪主要环境影响,结合环评现状,分析环评管理存在的主要问题,结合相关法规条文,查找存在这些问题的主要原因,结合目前核磁共振仪的建设情况和使用条件,从切合实际角度出发,提出核磁共振成像仪建设项目环评管理建议.     

核设施营运单位邻避问题应对策略研究

邻避问题激发的群体性事件是当前危险设施选址面临的主要问题,而核设施则是邻避设施中负外部性等级较高的一类设施.通过分析各行业邻避问题的九因素分析框架引入核设施邻避问题的研究,并结合国内外核行业和值得借鉴的非核行业典型案例进行分析,发现了影响邻避问题的核心要素包括程序公正、企业与政府公众形象、行业污名化倾向和补偿不合理等方面内容.核设施运营单位应在这些方面制定针对性策略和措施以减小邻避问题对核行业发展的影响.     

加载减速排放测试数据统计分析方法的改进研究

加载减速法（LugDown）测量被测车辆处于100％、90％和80％VelMaxHP三个速度工况下的排气光吸收系数（k）和100％VelMaxHP速度工况下的最大轮边功率值（MaxHP）,数据内涵增强,统计分析较为复杂。为解决此问题,本文根据LugDown法排气控制指标与测量数据特点,提出了排放最大值均值和功率比值统计分析方法,并将其用于广州市柴油车LugDown法排气检测数据的分析,结果令人满意。     

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。     

Alien plant invasions--incorporating emerging invaders in regional prioritization: a pragmatic approach for Southern Africa

Plant invasions are a serious threat to natural and semi-natural ecosystems worldwide. Most management-orientated research on invasions focuses on invaders that are already widespread and often have major impacts. This paper deals with "emerging" invaders-those alien species with the potential to become important problems without timely intervention. A climate matching procedure was developed to define areas of South Africa, Lesotho and Swaziland that could be invaded by 28 plant species that had previously been classified as emerging invaders. Information on the location of populations of these species in the study area was combined with information on their distributions (as native or alien) in parts of Australia and the United States of America. These two countries had the best available distribution data for this study. They also share many invasive alien plant species with South Africa. Climatic data obtained for weather stations near points of known occurrence in these countries were used to define the climatically suitable areas for each species in the study area. Almost 80% of the remaining natural environment in southern Africa was found to be vulnerable to invasion by at least one of these species, 50% by six or more and 24% by 16 or more species. The most vulnerable areas are the highveld grasslands and the eastern escarpment. The emerging invaders with the greatest potential range included Acacia podalyriifolia and Cortaderia selloana. The globally important invaders Ulex europaeus and Lythrum salicaria had a more limited invasion potential but could still become major invaders. There was no relationship between the extent of the climatically suitable areas for the different species and an expert ranking of their invasion potential, emphasising the uncertainties inherent in making expert assessments based on very little information. The methods used in this analysis establish a protocol for future modelling exercises to assess the invasion potential of other emerging invaders.