A multi-path routing-inventory problem for a closed-loop supply chain considering the heterogeneous fleet of vehicles

In this paper, the inventory-routing problem is studied for a closed-loop supply chain. This closed-loop supply chain considers suppliers, manufacturers, whole-sellers, and disposal centers. To formulate this problem, a mixed integer linear programming model is proposed. This mathematical model minimizes the total costs of the supply chain, including the fixed and variable costs of vehicles, and holding inventory costs of final products and scraps. The proposed model considers the road roughness degree, multi-path setting and the heterogeneous fleet of vehicles, which increases its flexibility and the quality of solutions. Then, two symmetry-breaking constraints are proposed to reduce the complexity of the mathematical model. In order to evaluate the integrity of the proposed model, 20 instances of different sizes are randomly generated and solved. Finally, a comprehensive sensitivity analysis is conducted with respect to five key features of the problem, such as the impact of the symmetry-breaking constraints on the CPU time, multi-path setting, fixed cost of vehicles, heterogeneous fleet of vehicles, and lost sales. The results indicate that the consideration of multi-path setting and the heterogeneous fleet of vehicles improves the quality of solutions significantly.     

碘化阶层孔氧化硅纳米球对水中典型有机氯污染物的吸附性能研究

采用共缩聚法,成功地原位合成了碘改性的阶层多孔氧化硅纳米球（SiO₂-I）,研究了其基于卤键作用对典型有机氯污染物六六六的吸附性能,并考察了改性剂I-硅烷含量和pH值对吸附效果的影响.实验结果表明,该纳米材料对六六六表现出优异的吸附富集性能,吸附速率快,60 min内对六六六的去除率为71.6%,240 min内达到吸附平衡,去除率可达98.3%,最大吸附量为178.6 mg·g^-1;吸附动力学符合拟二级动力学模型;碘物种的加入提高了阶层多孔氧化硅的吸附速率和吸附效率.另外,为了便于纳米吸附材料的分离,本研究对SiO₂-I纳米球进行了磁性化.研究发现,磁性化修饰后,SiO₂-I纳米材料仍然保留对六六六优异的吸附富集性能,240 min时的吸附去除率达90.3%.     

保定市大气污染特征和潜在输送源分析

保定市是京津冀地区重要城市之一.为了解保定市大气污染物质量浓度特征和潜在输送源,对保定市国控点2017年1月1日-12月31日PM₁₀、PM_2.5、SO₂、NO₂、O₃、CO等常规大气污染物数据进行分析,并利用TrajStat后向轨迹模型进行区域传输研究.结果表明:①ρ（PM₁₀）、ρ（PM_2.5）、ρ（SO₂）、ρ（NO₂）分别为（138±96）（84±66）（29±23）和（50±24）μg/m3,与2016年相比分别下降5.9%、9.1%、25.5%和13.1%;ρ（CO）较2016年下降了14.0%;ρ（O₃）较2016年增长了25.2%.ρ（PM₁₀）、ρ（PM_2.5）、ρ（NO₂）和ρ（O₃）分别超过GB 3095-2012《环境空气质量标准》二级标准限值的0.97、1.40、0.25和0.34倍,ρ（SO₂）和ρ（CO）未超标.②除ρ（O₃）外,其他污染物质量浓度均呈冬季最高、夏季最低的季节性特征,其中,冬季PM_2.5污染最为严重,春季PM_2.5~10（粗颗粒物）污染严重.③空气质量模型源解析结果显示,保定市ρ（PM_2.5）约60.0%~70.0%来自本地污染源排放.后向轨迹结果表明,在外来区域传输影响中,保定市主要受到西北方向气团（占比为21.7%~60.0%）远距离传输和正南方向气团（占比为34.8%~50.5%）近距离传输的影响.④PSCF（潜在源贡献因子分析法）和CWT（浓度权重轨迹分析法）分析表明,除保定市及周边区县本地污染贡献外,位于太行山东麓沿线西南传输通道的邯郸市、邢台市、石家庄市是影响保定市PM_2.5的主要潜在源区.研究显示,PM_2.5为保定市大气中的主要污染物,并呈冬季高、夏季低的变化特征,其主要来自西北远距离输送和南部近距离传输.      

人工智能技术对长江流域水污染治理的思考

随着经济的快速发展和城市化进程的不断加速,促使水污染严重的长江流域需从污染物去除过程的建模与优化、污水处理过程的优化控制、水污染监测系统的构建开展水污染治理研究.传统的水污染处理技术存在污染物去除效率预测精度较低、污水优化控制成本较高、水污染监测滞后效应严重的问题.人工智能技术能够有效克服上述问题,因此通过梳理国内外学者利用人工智能技术在污水污染物去除过程的建模与优化、污水处理过程的优化控制及水污染监测系统的构建等方面的研究成果,为全面加强长江流域水污染治理能力提供科学可靠的技术指导.结果表明:①利用人工神经网络技术（径向基神经网络、多层前馈网络-人工神经网络、多层感知器神经网络）对污水污染物去除过程进行建模与优化,为精确预测长江流域重金属（Cr、Cu）、营养盐（TN、TP）、持久性有机污染物〔PBDEs（多溴二苯醚）、HCH（六氯环己烷）〕的去除率提供重要参考价值.②采用污水处理的自动控制技术与人工智能技术（递归神经网络、支持向量机、模糊神经网络等）构建污水智能控制系统,为长江流域实现高效节能的污水优化控制提供重要的技术指导.③利用在线监测仪器和人工智能技术（小波神经网络、多元线性回归-人工神经网络、叠层去噪自动编码器等）建立水污染智能监测系统,为解决长江流域水污染监测响应滞后问题提供有力的技术支持.因此,人工智能技术对长江流域提高污水污染物去除率,降低污水优化控制成本,提升水污染监测时效性具有重要的推广价值.      

Pt负载暴露(001)晶面Bi2WO6催化剂合成及光催化性能

通过水热法制备了暴露（001）晶面的Bi₂WO₆纳米片,利用光还原法将Pt纳米颗粒负载于其表面.选择苯甲醇氧化和罗丹明B（RhB）降解为探针反应,评价了催化剂的光催化性能.在苯甲醇氧化实验中,Pt负载暴露001晶面的Bi₂WO₆样品的苯甲醇转化率为20.7%,约为未负载样品的2倍.在RhB降解实验中,Pt负载样品在光照40min后对RhB的矿化率可达81.1%,而未负载样品RhB矿化率仅为55.8%,表明Pt负载样品具有更优的降解速率和矿化能力.催化剂性能的提升归因于高能晶面暴露和Pt负载的协同作用.Pt纳米颗粒的负载作为助催化剂增加了催化剂表面的活性位点,同时提高了晶面光生电子空穴对的分离和迁移效率.     

基于高精度清单的京津冀机动车控污政策研究

为研究京津冀机动车污染控制政策对CO、HC、NO_x、PM_2.5、PM₁₀等污染物的减排效果,建立了2014年高精度机动车排放清单,选取过去已经实施,未来规划实施及优化/劣化后的若干政策,设置4类共13种政策情景,与实际清单基准情景进行对比,识别各政策情景的排放变化,并对其中由政府给予补贴的政策进行成本收益核算.结果显示,淘汰低排放标准机动车带来的污染减排效果最好,在天津和河北对CO和HC的排放削减分别为53.19%,49.75%和51.28%,50.87%,达半数及以上;升级机动车发动机和燃油标准也能显著削减排放,在天津和河北对PM_2.5、PM₁₀的削减分别为17.01%,17.00%和21.95%,21.93%.政府补贴政策存在明显边际效应特征,排放标准高,重污染车少的北京单位成本的减排收益明显低于天津和河北,.河北和天津在考虑成本因素的基础上,应当逐步采纳北京的高标准减排政策;北京则可在一定条件下,将一部分低效的政府补贴通过合理方式转移支付给天津和河北,以提高政府投入的减排效率.     

Analysis of pollutant levels in central Hong Kong applying neural network method with particle swarm optimization

Air pollution has emerged as an imminent issue in modernsociety. Prediction of pollutant levels is an importantresearch topic in atmospheric environment today. For fulfillingsuch prediction, the use of neural network (NN), and inparticular the multi-layer perceptrons, has presented to be acost-effective technique superior to traditional statisticalmethods. But their training, usually with back-propagation (BP)algorithm or other gradient algorithms, is often with certaindrawbacks, such as: 1) very slow convergence, and 2) easilygetting stuck in a local minimum. In this paper, a newlydeveloped method, particle swarm optimization (PSO) model, isadopted to train perceptrons, to predict pollutant levels, andas a result, a PSO-based neural network approach is presented. The approach is demonstrated to be feasible and effective bypredicting some real air-quality problems.     

固定化微生物处理有机污染物的研究进展

综述了近年来固定化微生物技术用于有机污染物治理中的最新进展，这些有机污染物包括难降解有机污染物（酚类、芳香烃类等）及其它化合物（DMP、甲硫醇恶臭气体、制药废水等）。此外，对藻类固定化技术作了简介，并对固定化技术的应用前景及存在问题进行了评述。     

Forest Area in Costa Rica: A Comparative Study of Tropical Forest Cover Estimates over Time

Forest area figures, at a given point in time and for a given region of interest, differ considerably, affecting the calculation of deforestation rates and thus confuse the political and scientific discussion on the state and change of the resource forest. This article discusses the variation of published forest cover figures, using Costa Rica as an example. A list of published figures on the forest cover of Costa Rica from 1940 onwards is analyzed. Reasons for the differences are hypothesized and discussed. These differences are mainly in the definition of forest and forest classes included, in the type of the studies conducted (mapping and/or sampling), in the precision of the estimates, and in the information sources used. It is concluded that part of the problem is inherent in the nature of the resource `forest'. Quality and completeness of the presentation of the forest cover estimates are a clue to their correct understanding and interpretation. The latter point being especially relevant, as forest cover data have both a technical-scientific and a political meaning and are used as relevant arguments in many discussions. In the example of Costa Rica, a general downward trend is observed up to about 1985/1990, whereas after that forest area figures are on the average at a markedly higher level. Some hypotheses for this change in the trend are discussed.     

A framework for assessment and monitoring of arthropods in a lowland tropical forest

By applying principles of adaptive management, and by using the valuable information that arthropods provide from assessment and monitoring programs, managers can identify and reduce possible impacts on biodiversity in development projects. In 1996, the Smithsonian Institution's Monitoring and Assessment of Biodiversity program worked together with Shell Prospecting and Development Peru to establish an adaptive management program to protect biodiversity in a natural gas exploration project in a Peruvian rainforest. In this paper, we outlined the conceptual steps involved in establishing an assessment and monitoring program for arthropods, including setting objectives, evaluating the results and making decisions. We also present the results of the assessment using some of groups of arthropods, and summarize the steps taken to identify appropriate groups for monitoring.