MSWNet: A visual deep machine learning method adopting transfer learning based upon ResNet 50 for municipal solid waste sorting

● MSWNet was proposed to classify municipal solid waste. ● Transfer learning could promote the performance of MSWNet. ● Cyclical learning rate was adopted to quickly tune hyperparameters. An intelligent and efficient methodology is needed owning to the continuous increase of global municipal solid waste (MSW). This is because the common methods of manual and semi-mechanical screenings not only consume large amount of manpower and material resources but also accelerate virus community transmission. As the categories of MSW are diverse considering their compositions, chemical reactions, and processing procedures, etc., resulting in low efficiencies in MSW sorting using the traditional methods. Deep machine learning can help MSW sorting becoming into a smarter and more efficient mode. This study for the first time applied MSWNet in MSW sorting, a ResNet-50 with transfer learning. The method of cyclical learning rate was taken to avoid blind finding, and tests were repeated until accidentally encountering a good value. Measures of visualization were also considered to make the MSWNet model more transparent and accountable. Results showed transfer learning enhanced the efficiency of training time (from 741 s to 598.5 s), and improved the accuracy of recognition performance (from 88.50% to 93.50%); MSWNet showed a better performance in MSW classsification in terms of sensitivity (93.50%), precision (93.40%), F1-score (93.40%), accuracy (93.50%) and AUC (92.00%). The findings of this study can be taken as a reference for building the model MSW classification by deep learning, quantifying a suitable learning rate, and changing the data from high dimensions to two dimensions.     

Environment and economic feasibility of municipal solid waste central sorting strategy: a case study in Beijing

Although Beijing has carried out municipal solid waste (MSW) source separation since 1996, it has largely been ineffective. In 2012, a “Green House” program was established as a new attempt for central sorting. In this study, the authors used material flow analysis (MFA) and cost benefit analysis (CBA) methods to investigate Green House’s environment and economic feasibility. Results showed that the program did have significant environmental benefits on waste reduction, which reduced the amount of waste by 34%. If the Green House program is implemented in a residential community with wet waste ratio of 66%, the proportion of waste reduction can reach 37%. However, the Green House is now running with a monthly loss of 1982 CNY. This is mainly because most of its benefits come from waste reduction (i.e., 5878 CNY per month), which does not turn a monetary benefit, but is instead distributed to the whole of society as positive environmental externalities. Lack of government involvement, small program scale, and technical/managerial deficiency are three main barriers of the Green House. We, thus, make three recommendations: involve government authority and financial support, expand the program scale to separate 91.4 tons of waste every month, and use more professional equipment/technologies. If the Green House program can successfully adopt these suggestions, 33.8 tons of waste can be reduced monthly, and it would be able to flip the loss into a profit worth 35034 CNY.

Open image in new window

     

上海市社区内影响居民垃圾分类效果的关键措施指标研究

本研究依托复旦大学意愿—习惯垃圾分类行为理论对上海市不同区域、不同类型的55个社区进行大范围调研,识别出包括设施、志愿者、保洁员、启动、定时定点、租户、管理和绿色账号在内的8个社区内影响居民垃圾分类效果的关键措施指标,并且通过指标与社区实际分类效果的相关性结果证明了该指标体系的有效性。研究结果还表明,上海市社区内影响垃圾分类效果的最主要的措施指标是硬件设施、保洁人员、“定时定点”制度和志愿者。研究结果一方面可以供上海市基层政府人员使用指标体系来进行监督,另一方面也为社区因地制宜地采取措施来提高和巩固社区垃圾分类效果提供了具体的科学支撑。     

Enrichment of fetal nucleated cells from maternal blood: Model test system using cord blood

The presence of small numbers of fetal nucleated red cells in the maternal circulation has been a stimulus for the development of technologies for non-invasive prenatal genetic analysis. Our laboratory has been assessing the feasibility of density gradient centrifugation followed by magnetic activated cell sorting (MACS) of cells expressing CD32 and CD45, to deplete maternal nucleated blood cells. We have examined the efficiency of each of the steps of this procedure using cord blood from term pregnancies as a source of nucleated red blood cells. Cord blood was shown to contain highly variable numbers of nucleated red cells. Three different density gradients were examined. There was no major difference in the performances of the double and triple gradients. Density gradient centrifugation resulted in enrichments of nucleated red blood cells of about 1000-fold relative to the total cell count. However, it was apparent that the selection of the cell layers which were most enriched for these cells would result in significant losses of nucleated red cells in other layers. MACS sorting of cells using CD45 resulted in white cell depletions ranging from 7 to 34-fold. These data provide a foundation for comparison with other methods and for optimization of the MACS technique.     

医疗垃圾的现状及处理对策

对医疗垃圾的危害及现有处理方法进行分析论述,指出医疗垃圾管理中存在的问题,在此基础上提出应重视医疗废物与生活垃圾的区分和对医疗废物的分类管理,医疗垃圾的最佳处置方法是热解气化技术,医疗废水的处理方法应建立在科学论证的基础上,医疗废物应该被合理处置、规范地科学地管理。     

垃圾分选设备及垃圾处理技术展望

我国是人口大国 ,随着人民生活水平的不断提高 ,现有的垃圾处理技术已逐渐不能适应产量日益增长与成分不断复杂化的垃圾处理的需要 ,垃圾处理急需一场革命性的变化 ,必须摆脱现有城市垃圾处理三大方法面临的技术上难以避免的弊端和“阴影”。也就是说 ,垃圾处理应该从垃圾分选设备技术整体创新入手 ,引入一整套的垃圾机械处理联合装置及各种新型的联合处理方法 ,使垃圾处理成本大大降低 ,从而产生一次历史转折性的突破 ,带动和实现垃圾处理的产业化。为了达到这一目标 ,对垃圾处理需要重点解决的技术难关进行了探讨与展望     

生活垃圾分类处理方式的生态效率评价

中国生活垃圾产量大、增速快，若处理方式不当会给环境和社会带来巨大危害，近年来提倡的垃圾分类处理方式是垃圾减量化的关键一环.综合考虑垃圾处理对环境、能源、经济和社会四方面影响，构建适用于垃圾处理的3E+S生态效率评价模型.将垃圾源头分类、末端分类处理方式纳入垃圾处理生命周期中构建4种生活垃圾处理情景.结果表明，混合收运+末端分类的处理情景具有潜在最优生态效率，分类收运+焚烧的处理情景由于源头分类、收运环节更为复杂导致其总环境影响潜值最大，且生命周期成本最高.对比考虑或不考虑社会成本计算得到的生态效率表明，社会成本的加入显著影响生态效率.源头分类是影响生活垃圾分类处理生态效率的敏感性因素，通过降低垃圾源头分类单元过程多余的环境排放和成本或推广使用以末端机械化分类（PGAS）技术为代表的低污染低成本的垃圾分类新技术，都能够提高垃圾处理的生态效率.     

混合多目标遗传算法求解地下水污染修复管理模型

为了提高多目标遗传算法Pareto解的局部最优性,本文将快速非支配遗传算法(NSGAII)与一种迭代式的局部搜索算法(Hill Climber with Step,HCS)相结合,开发了一种新的混合多目标遗传算法NSGAII-HCS.利用CONV1和ZDT6两个经典的多目标优化函数对NSGAII-HCS的性能进行测试,与传统的多目标算法NSGAII相比,CONV1得到的Pareto锋面与真实Pareto最优解锋面的平均距离由5.49减小到1.74,ZDT6则由0.16减小到0,表明NSGAII-HCS在保证解多样性的前提下,能使解接近或收敛到真实的Pareto最优解锋面.最后,将NSGAII-HCS与地下水流模拟软件MODFLOW和溶质运移模拟软件MT3DMS相耦合,并应用到一个理想的二维地下水污染修复管理模型中,结果分析表明该方法可为地下水污染治理提供多样的和收敛的Pareto管理策略,是一种稳定可靠的多目标优化方法.     

生活垃圾生物预处理后筛分及分选特性试验研究

堆肥化技术作为垃圾预处理方式在国内外已得到了广泛认可.本文通过现场试验,研究了生物预处理对垃圾筛分和分选特性的影响.结果表明,预处理后垃圾含水率降低了7.69%,堆体的体积从275 m~3减少到183 m~3,垃圾的容重从最初的356 kg·m~(-3)增加到470 kg·m~(-3).经过生物预处理,有机物不断被降解.预处理可以提高筛分后筛下物比例,预处理后的筛下物百分比提高了17%.预处理对可回收成分(金属、塑料、纸张、玻璃和布匹等)分选效率都较新鲜垃圾有不同程度的提高,对金属分选效果提高最显著,达到22.01%.本研究表明,生物预处理对垃圾筛分和分选均具有积极的作用,可以提高后期资源回收利用率.     

聚类问题集合论解法初探

本文主要探讨了应用集合论来解聚类问题的方法即集合聚类法,它是在论域上寻找由等价关系所诱导的划分。