2018年国家级空气质量主客观预报TS评分对比检验

目前对空气质量主客观预报效果的对比检验较少,为了解国家级空气质量主客观预报性能,利用预兆评分(threat score,TS)检验了2018年中央气象台主客观预报效果。结果表明:2018年,主(客)观预报对全国轻度及以上污染TS评分、空报率和漏报率分别为0.23~0.34(0.24~0.26)、0.37~0.43(0.39~0.41)和0.58~0.72(0.68~0.71)。说明主观预报整体上优于客观预报,但客观预报能力已接近主观预报。主客观预报的评分均在污染较重地区(京津冀及周边、汾渭平原、华中和长江三角洲地区)高于污染较轻地区(西北、西南、珠江三角洲和东北地区),在污染较重的冬季高于清洁的夏季。随着预报时效的延长,主观预报TS评分呈下降趋势,但客观预报TS评分变化不大。在污染较重的冬季,48和72 h时效客观预报TS评分高于主观预报。此外,在2018年5次重污染天气过程中,主(客)观预报对轻度、中度和重度污染的TS评分分别为0.39~0.57(0.43~0.46)、0.22~0.46(0.25~0.30)和0.10~0.34(0.10~0.18),主观预报TS评分高于客观预报的情况占了3次。随着预报时效的延长,客观预报表现更加稳定,说明客观预报能在污染过程中为预报员提供稳定参考,但对污染高值的预报能力仍需提高。     

基于核校准和SVM的煤矿安全组织管理因素分析

组织管理因素与煤矿安全生产密切相关。该文从人的管理、组织机构管理、企业环境管理、现场及技术管理四个方面详细分析了煤矿组织管理中的主要影响因素,提出一种基于核校准和支持向量机（SVM）的组织管理安全评价等级预测方法,构建了影响因素形成的输入空间到评价等级空间的非线性映射,并详尽分析了组织管理的安全评价等级与各影响因素的关联关系。在此基础上,本文深入研究煤矿安全中各组织管理因素的作用规律,并探讨了相应的预防措施与控制方法。实例分析结果表明,该方法能够合理地反映煤矿企业的组织管理安全评价状况,有助于煤矿企业采取相应措施预防或减少安全事故的发生,同时,也为企业提高本质安全管理水平提供了参考,有较强的实用性。     

基于支持向量机的白云湖水质评价

利用支持向量机模型良好的推广和泛化能力,以及在处理分类问题及小样本问题方面的优势,构建了湖泊水质评价模型,并运用此模型对白云湖的水质评价进行了研究。于2011年对广州市白云湖进水口A和出水口E以及湖区内B、C、D共5点分别进行了3次水质监测(1月份、4月份和8月份)。分析结果表明,A、B、C、D、E 5点1月份的水质除B点为Ⅳ类水外其余4点均为Ⅴ类水,4月份除A点为Ⅴ类水外其余4点均为Ⅳ类水,8月份除A为Ⅳ类水外其余4点均为Ⅱ类水,白云湖在经过开始的不稳定状态后,正在逐渐实现其净化水质的设计作用。相对于常规的评价方法,所得结果更为科学、合理。     

A fuzzy analytic hierarchy process (FAHP) approach to eco-environmental vulnerability assessment for the danjiangkou reservoir area,China

The Danjiangkou reservoir lies in the upper Hanjiang basin and is the source of water for the Middle Route Project (MRP) under the South-to-North Water Transfer Scheme (SNWT) in China. The eco-environment of water resource areas plays an important role in water conservation and purification and would have significant implications for the economic prosperity in Hanjiang basin as well as for the SNWT. Focusing on the Danjiangkou Reservoir Area (DRA), this study established an environmental information system database. Based on the database, an eco-environmental vulnerability assessment method using integrated fuzzy AHP (FAHP) and GIS was developed for the DRA. According to eco-environmental conditions and anthropic effects, vulnerability was classified into five levels: potential, light, medium, heavy and very heavy. The eco-environmental vulnerability distribution and major problems of the eco-environment were analysed and discussed. The results indicated that eco-environmental vulnerability in the DRA was moderate overall. Regions with lower eco-environmental vulnerability were located in Qinling Mountain area in the northwest, Daba Mountain area in the south and the area immediately surrounding Danjiangkou Reservoir in the east. Two regions with very high eco-environmental vulnerability were located in the north of Danjiangkou Reservoir in Henan province and in the western part of Shanxi province. On the basis of analysis of distribution of the different factors of eco-environmental vulnerability, different environmental protection measures were proposed for different areas. This study demonstrated that the proposed method was an effective approach for the assessment of environmental vulnerability. The results gained closely reflected the reality of the eco-environmental vulnerability of the DRA.     

In silico prediction of Tetrahymena pyriformis toxicity for diverse industrial chemicals with substructure pattern recognition and machine learning methods

There is an increasing need for the rapid safety assessment of chemicals by both industries and regulatory agencies throughout the world. In silico techniques are practical alternatives in the environmental hazard assessment. It is especially true to address the persistence, bioaccumulative and toxicity potentials of organic chemicals. Tetrahymena pyriformis toxicity is often used as a toxic endpoint. In this study, 1571 diverse unique chemicals were collected from the literature and composed of the largest diverse data set for T. pyriformis toxicity. Classification predictive models of T. pyriformis toxicity were developed by substructure pattern recognition and different machine learning methods, including support vector machine (SVM), C4.5 decision tree, k-nearest neighbors and random forest. The results of a 5-fold cross-validation showed that the SVM method performed better than other algorithms. The overall predictive accuracies of the SVM classification model with radial basis functions kernel was 92.2% for the 5-fold cross-validation and 92.6% for the external validation set, respectively. Furthermore, several representative substructure patterns for characterizing T. pyriformis toxicity were also identified via the information gain analysis methods.     

Fair fund distribution for a municipal incinerator using GIS-based fuzzy analytic hierarchy process

Burning municipal solid waste (MSW) can generate energy and reduce the waste volume, which delivers benefits to society through resources conservation. But current practices by society are not sustainable because the associated environmental impacts of waste incineration on urbanized regions have been a long-standing concern in local communities. Public reluctance with regard to accepting the incinerators as typical utilities often results in an intensive debate concerning how much welfare is lost for those residents living in the vicinity of those incinerators. As the measure of welfare change with respect to environmental quality constraints nearby these incinerators remains critical, new arguments related to how to allocate the fair fund among affected communities became a focal point in environmental management. Given the fact that most County fair fund rules allow a great deal of flexibility for redistribution, little is known about what type of methodology may be a good fit to determine the distribution of such a fair fund under uncertainty. This paper purports to demonstrate a system-based approach that helps any fair fund distribution, which is made with respect to residents' possible claim for fair damages due to the installation of a new incinerator. Holding a case study using integrated geographic information system (GIS) and fuzzy analytic hierarchy process (FAHP) for finding out the most appropriate distribution strategy between two neighboring towns in Taipei County, Taiwan demonstrates the application potential. Participants in determining the use of a fair fund also follow a highly democratic procedure where all stakeholders involved eventually express a high level of satisfaction with the results facilitating the final decision making process. It ensures that plans for the distribution of such a fair fund were carefully thought out and justified with a multi-faceted nature that covers political, socio-economic, technical, environmental, public health, and industrial aspects.     

河流水质评价中模糊数学评价法的应用与比较

选择2005年6个典型日（1月12日、3月10目、5月10日、7月13日、9月8日和11月10日）的水质,运用两种模糊数学评价法对浙江西苕溪流域港口断面的水质进行评价并作比较。第一种数学评价法采用较为简单的单因素模糊综合评价法,另一种采用多级模糊数学评价法。评价结果表明,1月12日、3月10日和11月10日的水质达到国家Ⅰ类水质标准;而5月10目的水质相对最差。比较两种评价方法,单因素模糊综合评价法对单个因素的影响敏感,特别是超标值对最后的相对隶属度贡献大;而多级模糊评价法综合各个因素来评价现状水质,因而反映出水质的综合情况。结合港口断面的特点分析水质变差的原因,结论是：流域农业劳作带来的面源污染物质进入港口断面使得该断面在农作高峰期水质较差;同时,断面的水资源短缺问题较为严重,导致水环境容量较小,污染物较难稀释扩散。     

基于支持向量机(SVM)的民用飞机重着陆智能诊断研究

针对国内航空公司对于重着陆的判断方法存在的不足,提出采用支持向量机(SVM)建立重着陆的智能诊断模型;分析对重着陆产生影响的相关因素,在力学基础上揭示了重着陆的产生原理;利用快速存取记录器中记录的多个飞行参数的信息,采用B737机型的实际样本数据进行训练和验证。结果表明:该方法能有效判断出是否发生重着陆,其准确率高达92.86%,证明该重着陆智能诊断方法具有较强实际应用价值,为后续研究奠定了基础。     

用于早期火灾分类的非线性决策树支持向量机

对早期火灾信息进行研究,提出了一种基于非线性决策树的支持向量机多类分类模型。该模型利用非线性映射将样本投影到高维特征空间,比较每类样本在高维空间的分布情况,进行聚类构造出一个二叉决策树,使容易区分的类别从根节点逐层分类出来,有效克服了错分积累和避免不可分情况;同时,各个节点采用二值最小二乘小波支持向量机,以获得较高的泛化能力。该文将该模型用于早期火灾分类,并与BP神经网络、K近邻法和决策树方法进行比较,实验结果表明,该模型对早期火灾的识别率更高。     

城市土地生态利用水平测度及障碍因子诊断

从土地代谢理念出发,构建了城市土地生态利用水平的测评指标体系。根据该指标体系,以"加权和与加权积的混合算法"和"因子障碍度测算"为模型基础,对浙江省11个地级市的土地生态利用水平进行了综合测评。研究结果表明:①浙江省11个市的土地生态利用水平存在明显的空间差异,整体上浙东北部较高,西南部偏低;②城市土地生态利用水平与城市经济社会发展相关,但并不意味经济社会越发达,城市土地生态利用水平就会越高;③土地污染物的高排放是影响浙江省城市土地生态利用的最大障碍因子,其中单位面积废水排放影响最大;④对城市土地生态利用水平的测度根本目的不在于区分排名的先后,而在于找到差距和改善的方向。通过因子障碍度检验,发现构建科学合理的障碍因子管理体系,大力发展循环经济和推行清洁生产,开展土地生态文明教育是提高城市土地生态利用水平的重要路径选择。