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.     

大数据解析技术在大气环境监测中的应用研究

针对近年来新兴的大数据及挖掘、分析技术,对大数据解析技术在环境科学研究中的应用进行了分析和探析。以城市局部地区大气细颗粒污染物PM2.5浓度计算为例,筛选了PM2.5浓度历史数据、气象条件、交通状况、人群活动情况、网格道路状况等数据为影响特征量,分析了用神经网络法进行大数据解析和推演的原理和数学过程,是用数学、计算机、统计等方法研究解决多元、非线性复杂环境问题的一种新的探索。     

A method to estimate the impact of clear-cutting on nutrient concentrations in boreal headwater streams

Large-scale forestry operations, like clear-cutting, may impair surface water quality if not done with environmental considerations in mind. Catchment and country level estimates of nutrient loads from forestry are generally based on specific export values, i.e., changes in annual exports due to the implemented forestry operations expressed in kg ha⁻¹. We introduce here a specific concentration approach as a method to estimate the impact of clear-cutting on nutrient concentrations and export in headwater streams. This new method is potentially a more dynamic and flexible tool to estimate nutrient loads caused by forestry, because variation in annual runoff can be taken into account in load assessments. We combined water quality data from eight boreal headwater catchment pairs located in Finland and Sweden, where the effect of clear-cutting on stream water quality has been studied experimentally. Statistically significant specific concentration values could be produced for total nitrogen, nitrate, ammonium, and phosphate. The significant increases in the concentrations of these nutrients occurred between 2 and 6 years after clear-cutting. Significant specific concentration values could not be produced for total phosphorus and total organic carbon with the whole dataset, although in some single studies significant increases in their concentrations after clear-cutting were observed. The presented method enables taking into account variation in runoff, temporal dynamics of effects, and the proportional size of the treated area in load calculations. The number of existing studies considering large site-specific variation in responses to clear-cutting is small, and therefore further empirical studies are needed to improve predictive capabilities of the specific concentration values.     

聚硅酸钛制备及其助凝特性

采用在活化硅酸制备过程中加入适量Ti Cl4的方法,试制了一种新型无机高分子助凝剂—聚硅酸钛。研究了不同[Si]/[Ti]条件下制备的聚硅酸钛在固定p H值的水中分散物的粒径和Zeta电位随着活化时间的变化趋势,并与活化硅酸作比较。结果表明,在Si/Ti=20,活化48 h的条件下能制备出较为稳定的聚硅酸钛。与活化硅酸相比,聚硅酸钛在水中的分散物具有较大的粒径和更高的Zeta电位。通过对比制备的活化硅酸和聚硅酸钛的红外图谱发现,聚硅酸钛中有TiO-Si键,证明钛离子参与了聚合反应,新生成的化学键提高了它的聚合度和稳定性。采用硫酸铝为混凝剂,聚硅酸钛为助凝剂进行混凝除浊实验。测定了絮凝体的Zeta电位和混凝沉淀后的上清液浊度,研究了硫酸铝投加量和混凝p H对聚硅酸钛助凝效果的影响。结果表明,投加适量的聚硅酸钛助凝剂,明显提高了混凝除浊效果且拓宽了有效混凝p H值的范围。     

高氯低COD值水样的无汞快速测定法

以硝酸银和硫酸铬钾代替硫酸汞来消除 COD 测定中的氯离子干扰，同时将重铬酸钾溶液的浓度降低为0.100 mol／L，并用硫磷混酸代替硫酸，该方法经过对标准样品和实际样品的测定，表现出对于测定氯离子含量低于25000 mg／L的高氯低 COD 值水样具有较好的准确度和精密度，同时可以实现银盐的回收再利用。     

超高强度钻杆低温断裂性能及裂纹扩展特征研究

超高强度钻杆在冬季低温环境下容易发生脆性断裂事故,为了从断裂力学角度揭示其冷脆现象的本质,采用ZBC2302-D型示波冲击试验机,获得了不同温度下超高强度钢在冲击断裂过程中的力-位移曲线,并对比分析了冲击功、起裂功和裂纹扩展功之间的变化关系。采用国家标准GB 4161—2007中的计算公式,计算了冲击试样的裂纹扩展量Δa和阻抗应力强度因子K_R;同时,引入K_R阻力曲线,得到超高强度钻杆在不同温度下的剩余强度图。结果表明,当冲击试样开始出现裂纹时,随温度降低,钻杆的初始起裂应力有所增加,出现硬化现象。当钻杆出现一定长度的裂纹时,随温度降低,阻抗应力强度因子越来越小,而且降低速度越来越快,剩余强度也表现出相同的变化趋势。随温度降低,材料抵抗裂纹扩展的能力逐渐下降,表明钻杆在低温下更容易断裂。     

农药混配制剂环境风险评估现状与展望

本文综述了欧洲和美国农药混配制剂的环境风险评估方法。详细介绍了欧洲食品安全局(EFSA)评估体系中的2种方法,即,基础的"整体测试法"和近年来提倡的"基于组分的方法"。"基于组分的方法"的特点是以浓度加和模型(CA模型)作为默认假设进行初级评估,以独立作用模型(IA模型)等作为高级评估手段的农药混配制剂环境风险评估方法。此外,本文还介绍了模型偏差率(MDR)、毒性相似度及毒力单元(TU)等概念以及混配制剂风险评估流程。本文的目的旨在为建立我国农药混配制剂的环境风险评估方法体系提供参考。     

焊点热疲劳失效原因分析

某PCBA样品在使用约半年后出现功能失效,该PCBA在封装后进行整体灌胶,将失效样品剥离,发现部分器件直接脱落,通过表面观察、切片分析、EBSD分析、应力分析、热膨胀系数测试等手段对样品进行分析,结果表明:各封装材料存在热失配,且焊点缺陷较多且存在应力集中区,加速焊点的疲劳失效进程,导致PCBA功能失效。     

Fe~(3+)和Cu~(2+)对饮用水氯消毒副产物三卤甲烷形成的影响

为了研究水中存在Br-的情况下,Fe3+和Cu2+对三卤甲烷(THMs)的生成及CHCl3、CHBr Cl2、CHBr2Cl、CHBr34种消毒副产物相对分布的影响,以腐殖酸模拟氯消毒过程中的前体物进行实验。结果表明,在p H为6、7和9 3种条件下,Fe3+抑制了THMs的生成,p H=6时只有CHCl3生成量随着Fe3+浓度的增加逐渐减少,其余3种消毒副产物均在增加,p H=7时4种消毒副产物浓度均减小并在Fe3+浓度为2 mg/L时生成量最低,p H=9时的生成趋势与p H=6时类似。Cu2+能促进THMs的生成,在p H为6、7和9时,当加入0.5 mg/L Cu2+时,THMs总量分别增加了16.7%、22.6%和2.5%,随着p H增加,THMs总量增加。在3种p H环境中,Cu2+对THMs生成的影响大于Fe3+,在偏酸性环境中,Fe3+影响THMs生成,产生的致癌风险高,当金属离子浓度为2.5 mg/L时,致癌风险相差最高为15%,在中性和偏碱性环境中,Cu2+影响THMs生成,产生的致癌风险高。     

基于小波变换与传统时间序列模型的臭氧浓度多步预测

采用最大重叠小波分解与重构方法,将影响O3小时浓度的不同时间尺度的物化过程分离出来,以提高序列的光滑性.同时,选择合适的传统时间序列模型(如ARIMA模型等)来描述不同过程的序列特征,并分别拟合预报.最后,在建模中引入24 h季节项,以实现提前24 h-次性预测未来1d的O3逐时浓度.结果表明,预报的平均相对误差为12.92％,平均绝对误差和均方根误差分别为10.04 μ.g·m-3和13.98μg·m-3,预报值与实测值的相关系数和匹配指数分别为0.96和0.98.随着预测期的延长,预报误差仍处于可接受范围内.该方法同样适用于每日最大O3小时浓度预报,研究结果为发布天气预报式的空气质量预报提供了新思路,便于公众规划出行并减少大气污染暴露.