针对细微特征进行K-means聚类的电台分选识别技术

现代战场中的无线通信设备日益增多，精准获取个体信息已成为研究热点，但也是难点。针对通信电台，提出了一种分选识别技术。该技术从电台物理层特性出发，对其辐射信号的细微特征进行K-means聚类以实现分选，分选的同时提取各个个体的特征属性值，未知信号通过与特征属性值相关运算实现个体识别。该技术无需先验知识，无需训练运算，通过实验验证，其可行、高效，易于工程实现。     

Single Cell Bioprinting with Ultrashort Laser Pulses

Tissue engineering requires the precise positioning of mammalian cells and biomaterials on substrate surfaces or in preprocessed scaffolds. Although the development of 2D and 3D bioprinting technologies has made substantial progress in recent years, precise, cell-friendly, easy to use, and fast technologies for selecting and positioning mammalian cells with single cell precision are still in need. A new laser-based bioprinting approach is therefore presented, which allows the selection of individual cells from complex cell mixtures based on morphology or fluorescence and their transfer onto a 2D target substrate or a preprocessed 3D scaffold with single cell precision and high cell viability (93–99% cell survival, depending on cell type and substrate). In addition to precise cell positioning, this approach can also be used for the generation of 3D structures by transferring and depositing multiple hydrogel droplets. By further automating and combining this approach with other 3D printing technologies, such as two-photon stereolithography, it has a high potential of becoming a fast and versatile technology for the 2D and 3D bioprinting of mammalian cells with single cell resolution.     

干扰床分选机分选磷矿规律的试验研究

文中分析了我国丰富的磷矿资源及传统的分选磷矿方法,阐述了研究干扰床分选机分选磷矿的规律具有重要意义,研究设计了试验室干扰床分选机分选系统,试验采用宜昌某磷矿原矿中0~3.0mm粒度的粗矿泥进行试验研究,分析了干扰床分选机分选磷矿的规律,找出了分选最佳工况点并进行了对比曲线分析,论证了显著的经济效益和广阔的市场前景。     

工程建设标准客观优先度评价研究

在工程建设标准的制定和修订过程中，如何科学地确定编制顺序一直是困扰标准管理部门的难题，采用传统方式“按需制标”，具有盲目性、主观性，容易造成标准在制定和修订过程中的互相重复、滞后与矛盾。为了保证标准制定和修订工作的协调性，指导标准体系中各标准的制修订顺序，引入复杂网络拓扑结构，将同一标准类型下的标准构成功能团，根据相互之间的关系形成复杂网络，采用剥落算法对其网络节点重要性排序，进而确定标准制修订的优先度。通过对建筑给水排水工程规范标准修订的实证分析，总结工程建设标准的网络拓扑优先度评价的算法模型，实现制定和修订顺序的自动计算，从而更加科学地指导标准的制定和修订工作     

Measuring labor input on pasture-based dairy farms using a smartphone

With the cessation of milk quotas in the European Union, dairy herd sizes increased in some countries, including Ireland, with an associated increase in labor requirement. Second to feed costs, labor has been identified as one of the highest costs on pasture-based dairy farms. Compared with other European Union countries, Ireland has historically had low milk production per labor unit; thus, optimization of labor efficiency on farm should be addressed before or concurrently with herd expansion. The objective of this study was to quantify current levels of labor input and labor efficiency on commercial pasture-based dairy farms and to identify the facilities and management practices associated with increased labor efficiency. Thirty-eight dairy farms of varying herd sizes, previously identified as labor-efficient farms, were enrolled on the study and data were collected over 3 consecutive days each month over a 12-mo period, starting in May 2015 and finishing in August of 2016. This was achieved through the use of a smartphone application. For analysis purposes, farms were categorized into 1 of 3 herd size categories (HSC): farms with <150 cows (HSC 1), 150–249 cows (HSC 2), or ≥250 cows (HSC 3). Overall farm labor input increased with HSC with 3,015, 4,499, and 6,023 h worked on HSC 1, 2, and 3, respectively. A higher proportion of work was carried out by hired staff as herd size increased. Labor efficiency was measured as total hours input to the dairy enterprise divided by herd size. Labor efficiency improved as herd size increased above 250 cows with 17.3 h/cow per yr observed for HSC 3; labor efficiency was similar for HSC 1 and 2, at 23.8 and 23.3 h/cow per yr, respectively. A large range of efficiency was observed within HSC. The labor requirements had a distinct seasonal pattern across the 3 HSC with the highest input observed in springtime (February to April) primarily due to calving and calf-care duties, milking, and winter feeding. The lowest input was observed in wintertime (November to January) when cows were dry. Particular facilities and management practices were associated with efficiency within certain tasks, the most notable in regard to milking and winter feeding practices. Additionally, the most efficient farms used contractors to perform a higher proportion of machinery work on farm than the least efficient farms.     

关于构建塑料包装废弃物的分类、收集与管理体系的思考

简介了我国塑料包装行业发展状况和塑料包装废弃物治理的有关政策与法规；阐述了为解决塑料包装废弃物等对环境造成的污染不同主体应承担与落实的责任，包括政府部门、企业和消费者等；提出了为高值（质）化回收利用塑料包装废弃物构建细化分类、收集与管理体系的一些设想。     

‘I am not a number’: Conceptualising identity in digital surveillance

Surveillance, now a commonplace phenomenon in everyday life, has been explored from various disciplines over three decades. Today's surveillance practices depend primarily upon many software technologies that collect, store and process personal data for the purposes of influence, management, protection or detection. The identification and categorisation of data have thus emerged as the technical signature of surveillance. An individual has many identities belonging to different contexts of his/her life, but in this paper, we explore the relationship between surveillance and identity in virtual contexts only. We argue that an understanding of identity purely as data is fundamental to understanding surveillance. We propose abstract general definitions of surveillance and identity that together create a conceptual framework, capturing key features common to many disparate surveillance situations. Our work concludes that the essence of surveillance is that of a surveillance context, which is precisely and solely defined by the availability of data about the behaviour and identity of its entities. The data that distinguishes the entities of the context we call identifiers; we explore the creation, provenance, comparison and transformation of identifiers. Abstractly, surveillance is a process that tests for properties of data, and sorts identifiers into categories.     

结合深度学习与相关反馈的遥感图像检索

目的 针对基于内容的图像检索存在低层视觉特征与用户对图像理解的高层语义不一致、图像检索的精度较低以及传统的分类方法准确度低等问题，提出一种基于卷积神经网络和相关反馈支持向量机的遥感图像检索方法。方法 通过对比度受限直方图均衡化算法对遥感图像进行预处理，限制遥感图像噪声的放大，采用自学习能力良好的卷积神经网络对遥感图像进行多层神经网络的监督学习提取丰富的图像特征，并将支持向量机作为基分类器，根据测试样本数据到分类超平面的距离进行排序得到检索结果，最后采用相关反馈策略对检索结果进行重新调整。结果 在UC Merced Land-Use遥感图像数据集上进行图像检索实验，在mAP（mean average precision）精度指标上，当检索返回图像数为100时，本文方法比LSH（locality sensitive Hashing）方法提高了29.4%，比DSH（density sensitive Hashing）方法提高了37.2%，比EMR（efficient manifold ranking）方法提高了68.8%，比未添加反馈和训练集筛选的SVM（support vector machine）方法提高了3.5%，对于平均检索速度，本文方法比对比方法中mAP精度最高的方法提高了4倍，针对复杂的遥感图像数据，本文方法的检索效果较其他方法表现出色。结论 本文提出了一种以距离评价标准为核心的反馈策略，以提高检索精度，并采用多距离结合的Top-k排序方法合理筛选训练集，以提高检索速度，本文方法可以广泛应用于人脸识别和目标跟踪等领域，对提升检索性能具有重要意义。     

Characterization of Extra Virgin Olive Oil from Southern Brazil

Six olive oils extracted from the cultivars Arbequina, Arbosana, Coratina, Frantoio, Koroneiki, and Picual from 2017 and 2018 harvests, cultivated in Pinheiro Machado, Rio Grande do Sul, Brazil, are evaluated for standard oil composition parameters and bioactive constituents (pigments, tocopherols, and phenolic compounds). Multivariate principal component analysis (PCA) and univariate ANOVA and Fisher's LSD test are used to verify the effect of cultivar and harvest year on oil composition. Olive oil composition met extra virgin olive oil (EVOO) standard parameters and is influenced by both cultivar and harvest year. EVOO produced in 2018 has greater chlorophyll, caffeic acid, ligstroside aglycone, hydroxyoleuropein aglycone, syringic acid, and hydroxytyrosol acetate contents than the EVOOs from 2017. Linoleic acid, ferulic acid, ligstroside aglycone, and hydroxytyrosol acetate are the variables whose contents most contributed to the differentiation of oils by cultivar in both harvest years. Chemical characterization analyses allow for the differentiation of oil composition based on harvest year and cultivar. Metabolic quality data obtained here support the establishment of a local EVOO profile and the compounds that most contributed to treatment differentiation may serve as markers that can be utilized in determining origin, cultivar, and harvest year. Practical Applications: Olive production in Brazil is recent and is based on European cultivars which have not been bred for the local environmental conditions. Therefore, the measurement of olive oil metabolic quality will determine cultivar adaptability to local edaphoclimatic conditions as well as assist in the establishment of a standard of identity for the product and promote the development of its market. Olive oil produced in Southern Brazil shows high quality, and is especially rich in phenolic compounds. Although harvest year influences oil composition, oil from both harvests meet EVOO standards and cultivar specific metabolic markers are observed. This study provides the foundation for olive producers in Southern Brazil to seek authentication of the geographical origin of olive oil.     

近50年大黑河流域气候水文要素变化趋势及突变分析

基于大黑河流域6个气象站1966～2013年气温、降水量、蒸发皿蒸发量、相对湿度和日照时数长期数据系列,采用气候倾向率和滑动t检验法分析了近50年大黑河流域气象要素的变化趋势及变异年份。结果表明,大黑河流域整体呈现暖干气候倾向,气温呈显著增高趋势,变化率为0.43～0.57℃/10a,气温增温变点大致出现了两次,分别在1986、1997年;降水量和蒸发量多呈不显著减少趋势,降水量先减(1981年)后增(1990年),蒸发皿蒸发量先减(1986年)后增(1996年);相对湿度、日照时数多呈显著减小趋势。