基于指数平滑技术的装备维修备件需求预测

以装备维修备件的需求量预测问题为研究对象,运用指数平滑预测技术,构建了装备维修备件需求量预测模型,并借助Excel环境对某装备维修备件的需求量进行了预测。预测结果表明,该方法为科学预测装备维修备件的需求量提供了一种可行的方法。     

支持向量机在环境空气质量评价中的应用

环境质量评价是环境科学的重要组成部分。建立环境空气质量评价的支持向量机模型,并用实例进行验证,评价结果与实际情况及其它方法相吻合,表明支持向量机在环境空气质量评价中的可行性和有效性。     

Development of marginal emission factors for N losses from agricultural soils with the DNDC–CAPRI meta-model

The article discusses marginal emission factors for N losses from agricultural soils, with rape and wheat as examples, and presents results for EU15 as high-resolution maps and aggregated to Member State level. The results are generated by linking the economic model for the agricultural sector CAPRI (Common Agricultural Policy Regional Impact) with spatial down-scaling, and a statistical meta-model for the bio-physical model DNDC (DeNitrification–DeComposition). For a given agro-economic scenario, CAPRI supplies for each crop the crop share, yield and fertilizer application rate spatially downscaled to clusters of 1 km × 1 km grid cells. The results from CAPRI are processed by a meta-model of DNDC to estimate the local greenhouse gas emissions from the soil. DNDC is a dynamic process-oriented model, which estimates trace gas fluxes and nutrient turnover in agricultural soils. The fit of the regressions is typically very good (0.95R² for the majority of the regressions), and all coefficients are significant at 99% probability. The meta-model allows a seamless integration between the economic and the bio-physical models, offering additional benefit such as the site-specific calibration of the bio-physical model ensuring the match between simulated and observed yield at the grid-level.The meta-model is used to calculate marginal emission factors for a 1 kg ha⁻¹ increase of mineral N and manure fertilizer rates for rape and wheat, at different levels of fertilization. They show that for Western European farming practice, only a small fraction of extra nitrogen fertilizer would go into increased yields: most of it would be emitted to the environment. The largest spatial variability is observed for N₂O emissions. The derivation of marginal emission factors is just one of the many possible uses for the linked regionalized agro-economic and soil chemistry model, which exploits to a large extent both geo-referenced and regionally available statistical information at European scale.     

突发事件应急决策探讨

随着我国社会经济的发展,突发事件呈现急剧上升的趋势。通过分析总结突发事件应急决策的特点,从多个角度分析比较其与常规决策的差异,总结介绍了应急决策方法,构建了应急决策的流程模型,从突发事件应急决策支持系统的框架模型说明了软件支持、信息支持、辅助决策三个层次需要解决的关键问题,最后总结了应急决策评价的标准和方法等。     

基于数字流域的水质综合管理决策支持系统——以深圳石岩水库流域为例

鉴于城市化地区水环境问题的时空复杂性,单一的水环境管理措施已不能满足实际管理的需要,为改善水环境有必要采取综合管理措施。文章以数字流域为基础,将流域内水系数据、地形数据、土地利用数据、社会经济及人口数据等进行融合,构建了水质综合管理决策支持系统,着重探讨了综合模型集成、动态空间数据与模型及管理之间的耦合关系问题;以深圳市石岩水库流域为例进行实证研究,构建了石岩水库流域管理决策支持系统,评价了旱季截污、产业结构调整、雨季截流等单个及组合措施的水质改善效果。结果表明该系统可以模拟多种措施的综合作用、动态的更新数据与模拟结果、并提供可视化的空间决策信息,是复杂流域水质动态综合管理与空间决策支持的有效工具。     

废润滑油再生决策支持系统的研究

针对废润滑油再生过程中的工艺选择,将决策支持系统引入废润滑油再生领域。提出了废润滑油再生工艺决策支持系统的理论框架。构建出基于数据库、模型库的废润滑油再生工艺选择决策系统。用户可通过良好的人机对话界面,根据润滑油污染程度和再生要求,确定满足需要的废润滑油再生工艺。     

小波支持向量机在大气污染物浓度预测中的应用

用小波分解重构和支持向量机相结合的方法,建立大气污染物浓度预测模型。通过小波分解,将大气污染物浓度序列分解为不同频段的小波系数序列,再对各层的小波系数序列重构到原尺度上。利用相关分析的方法构建出低频小波系数a3和中频小波系数d3的支持向量机模型输入因子为前一天小波系数a3和7个气象因子;高频小波系数d2和d1以前三日的小波系数为输入因子,然后对各小波系数序列采用相应的支持向量机模型进行预测,各小波系数均使用ν-支持向量回归机(ν-SVR)算法和径向基函数,最后通过小波重构合成大气污染物浓度序列的最终预测结果。通过对大气SO2浓度预测实例证明,该大气污染物浓度预测模型具有推广能力较强、预测精度较高、训练速度快、便于建模等优点,具有良好的应用前景。     

基于空间数据挖掘的区域危险废物决策支持设计研究

区域危险废物环境管理与风险控制是涉及到危险废物特性、环境、经济和社会等多因素的复杂过程,对资源环境的可持续发展和人体健康都至关重要。空间数据挖掘技术在传统数据挖掘的基础上增加空间维尺度,能够实现从空间数据库和非空间数据库中提取危险废物产生、转移、处置和管理过程中隐含的空间模式与特征。文章在传统的危险废物环境管理和风险控制方法的基础上,构建了区域危险废物空间数据挖掘的指标体系,并设计了一种基于空间数据挖掘技术的区域危险废物决策支持系统框架。该框架由数据层、数据挖掘层和显示层组成,为区域危险废物管理和决策提供了直观的、科学的和高效的支持,提供了一种新颖可行的思路和方法提高危险废物的环境管理和风险控制水平,具有一定的理论和现实意义。     

Safety analysis and risk assessment of a Micro-GtL Plant

Laboratory hydrogen generators, medical oxygen, and micro-breweries are examples of modular and micro technologies that are commercial successes. Researchers, patients, and unskilled workers operate these facilities but more complex processes require highly qualified personnel to ensure they operate safely. Modular-micro processes in isolated locations meet economic objectives when operated remotely thereby minimizing labor costs. Mitigating the risk requires a comprehensive hazards analysis with advanced control systems particularly for explosive and toxic compounds. Here, we propose a method called Failure Mode Risk Decision (FMRD) to review the inherent hazards of a micro-refinery unit (MRU) that converts flared and wasted natural gas to long chain hydrocarbons. This approach combines the Process Flow Failure Mode (PFFM) methodology as a systematic and reliable technique with a novel numerical risk assessment to improve the analytical evaluation of hazardous conditions. The objective is to combine causes and consequences in a single metric, where scaled probability of evident causes and severity of consequences are used to derive a risk level measure. With the proposed metric, the magnitude of a potential hazard is directly correlated with the risk level. This mechanism identifies extra risk scenarios compared to the classical hazard analysis method and provides a straightforward comprehensive numerical assessment to represent the inherent and residual risks to facilitate justifying the hazardous scenarios. Accordingly, we design a safety loop and supply all the required facilities to remove the potential risks at the process plant. Not only the proposed methodology clarifies the risks of the MRU presented in this study, but can be extended to review the hazards of other chemical process plants.