基于孪生神经网络的时间序列相似性度量

在时间序列分类等数据挖掘工作中，不同数据集基于类别的相似性表现有明显不同，因此一个合理有效的相似性度量对数据挖掘非常关键。传统的欧氏距离、余弦距离和动态时间弯曲等方法仅针对数据自身进行相似度公式计算，忽略了不同数据集所包含的知识标注对于相似性度量的影响。为了解决这一问题，提出基于孪生神经网络（SNN）的时间序列相似性度量学习方法。该方法从样例标签的监督信息中学习数据之间的邻域关系，建立时间序列之间的高效距离度量。在UCR提供的时间序列数据集上进行的相似性度量和验证性分类实验的结果表明，与ED/DTW-1NN相比SNN在分类质量总体上有明显的提升。虽然基于动态时间弯曲（DTW）的1近邻（1NN）分类方法在部分数据上表现优于基于SNN的1NN分类方法，但在分类过程的相似度计算复杂度和速度上SNN优于DTW。可见所提方法能明显提高分类数据集相似性的度量效率，在高维、复杂的时间序列的数据分类上有不错的表现。     

Effects of Ni Content and Ball Milling Time on the Hydrogen Storage Thermodynamics and Kinetics Performances of La–Mg–Ni Ternary Alloys

The effects of Ni content and ball milling time on the hydrogen storage thermodynamics and kinetics performances of asmilled La_5Mg_(95-x)Ni_x(x = 5, 10, 15) ternary alloys have been investigated.The evolution of microstructure and phase of experimental alloys in the absorption/desorption process has been characterized by XRD, SEM and HRTEM.The hydrogen storage kinetics and thermodynamics performances and PCI curves have been tested using the Sievert apparatus.It is found that the rising of Ni content remarkably improves the hydrogen storage kinetic performance, but reduces hydrogen storage capacity.And with the increase in milling time, hydrogen desorption activation( E_a) value decreases firstly and then increases; the minimum value is 47.6 kJ/mol, and the corresponding milling time is 10 h for La_5Mg_(85)Ni_(10) alloy.As for the thermodynamics properties, the hydrogenation enthalpy(Δ H) and hydrogenation entropy(Δ S) both decrease firstly and then increase with the rising of Ni content and milling time.The composite La_5Mg_(85)Ni_(10)alloy milled for 10 h exhibits the best thermodynamics and kinetics performances, the lowest E_a of 47.6 kJ/mol, absorption of 5.4 wt.% within 5 min and desorption of 5.2 wt.% within 3 min at 360 ℃ and the lowest Δ H and Δ S of 72.1 kJ/mol and 123.2 J/mol/K.     

基于词法特征的恶意域名快速检测算法

针对互联网中恶意域名攻击事件频发，现有域名检测方法实时性不强的问题，提出一种基于词法特征的恶意域名快速检测算法。该算法根据恶意域名的特点，首先将所有待测域名按照长度进行正则化处理后赋予权值；然后利用聚类算法将待测域名划分成多个小组，并利用改进的堆排序算法按照组内权值总和计算各域名小组优先级，根据优先级降序依次计算各域名小组中每一域名与黑名单上域名之间的编辑距离；最后依据编辑距离值快速判定恶意域名。算法运行结果表明，基于词法特征的恶意域名快速检测算法与单一使用域名语义和单一使用域名词法的恶意域名检测算法相比，准确率分别提高1.7%与2.5%，检测速率分别提高13.9%与6.8%，具有更高的准确率和实时性。     

数据驱动的机床等待过程节能方法研究

为了降低机床等待过程中的能耗，提出了一种实时数据驱动的机床等待时间预测与节能控制方法。首先，建立了射频识别驱动的生产进度评估方法，并以生产进度数据作为输入，构建了基于堆栈降噪自编码的机床等待时间预测模型；其次，依据预测的机床等待时间，提出了机床状态切换方法，以降低机床能耗；最后，通过一个电梯零部件制造车间的案例分析，表明该方法的预测误差仅为4.1%，同时将机床等待过程能耗降低了57%，实现了制造车间的节能减排。     

Effect of processing method on vitamin profile,antioxidant properties and total phenolic content of coconut (Cocos nucifera L.) sugar syrup

Coconut (Cocos nucifera L.) sugar is a more nutritious alternative sugar source as compared to sugar palm (Borassus flabellifer) and sugarcane (Saccharum officinarum L.). This work was aimed to investigate the browning index (BI), vitamin profile and antioxidant properties of coconut sap sugar syrups, which were produced by different processing methods: rotary evaporation (RE), microwave evaporation (ME) and open-heat evaporation (OHE). The results obtained showed that coconut sugar syrup produced by RE-60 contained high antioxidant activities [DPPH (36.71%) and ABTS (34.84%), TPC (299.87 mg per 100 g sample) and FRAP (3.74 mm )]. These values were slightly lower than those of ME and OHE. Coconut sugar syrup (RE-60) also contained higher amounts of vitamin C (1587.27 mg L⁻¹), vitamin B1 (97.44 mg L⁻¹) and vitamin B3 (19.84 mg L⁻¹) compared with those of ME and OHE coconut sugar syrups. RE-60 was the best method to produce coconut sugar syrup in a shorter time with lower browning index and higher vitamin contents.     

Emergency egress for the elderly in care home fire situations

Practitioners are continuing to develop egress modelling software for the design of the built environment. These models require data about human behaviour and factors for calibration, validation and verification. This study aims to address the specific data and knowledge gap: emergency egress of the elderly. Such data are difficult to collect given privacy and consent concerns, with strong relationships generally being required between residences and researchers. Through the observation of nine fire drills at six Canadian long-term care (LTC) and retirement homes, specific evacuation actions and behaviour were observed for 37 staff members and information about the evacuation of 56 residents was collected. These drills demonstrated that emergency egress in LTC and retirement homes is highly staff dependent with 72% of residents recorded requiring full assistance at all stages of movement in evacuation, and that the type of announced/unannounced drill and level of resident care will affect the type of data collected. The development of travel speed and pre-movement is discussed subject to limitation with qualitative behavioural insights of residents that were observed. This study provides valuable methodological discussion on how to conduct behavioural studies in similar highly restricted research environments. Specific attention is given to understanding the considerations that must be made when using fire drills as data sources, and the impact that these can have on using such data for modelling. This study may inform the initial setup and programming of evacuation models from an actions and behavioural perspectives of staff members and residents.     

求解带时间窗车辆路径问题的改进型烟花算法

针对带时间窗车辆路径问题,为寻求组合优化问题最优解,构建总运输成本最小数学模型。由于烟花算法搜索半径不能自适应调整,算法后期易陷入局部最优,故对爆炸算子进行改进,使最优烟花搜索半径自适应调整,增强后期局部搜索能力;同时利用分布式信息共享机制避免算法早熟并引入变异火花增强种群多样性。利用标准测试集进行验证后,结果表明该算法在求解带时间窗车辆路径问题时不仅具有可行性和有效性,并且收敛速度快、搜索质量高。     

两相区配分时间对IQ&P钢组织与性能的影响

以含Cu低碳钢为研究对象,利用SEM、EPMA和拉伸试验研究了两相区配分时间对其组织演变、元素配分以及经IQ&P处理后力学性能的影响,并利用Dictra软件对元素配分行为进行了动力学计算。结果表明,IQ处理后试验钢中的块状马氏体形成于原铁素体区域;随两相区配分时间延长,粒状马氏体数量减少,板条状马氏体之间的间距减小。检测和计算结果的对比显示,C、Mn、Cu 3种元素的相对配分速率与计算结果一致,但实际配分速率低于各自的计算结果。随两相区配分时间延长,经IQ&P处理后试验钢的抗拉强度先增加后减小,而伸长率持续减小;在600 s时达到较好的强塑性匹配,强塑积为16 963.24 MPa·%。     

Reference databases,primer choice,and assay sensitivity for environmental metabarcoding: Lessons learnt from a re‐evaluation of an eDNA fish assessment in the Volga headwaters

Biodiversity monitoring via environmental DNA, particularly metabarcoding, is evolving into a powerful assessment tool for riverine systems. However, for metabarcoding to be fully integrated into standardized monitoring programmes, some current challenges concerning sampling design, laboratory workflow, and data analysis need to be overcome. Here, we review some of these major challenges and potential solutions. We further illustrate three potential pitfalls, namely the choice of suitable metabarcoding primers, the necessity of complete reference databases, and varying assay sensitivities, by a reappraisal of our‐own recently carried out metabarcoding study in the Volga headwaters. TaqMan qPCRs had detected catfish (Silurus glanis) and European eel (Anguilla anguilla), whereas metabarcoding had not, in the same samples. Furthermore, after extending the genetic reference database by 12 additional species and re‐analysing the metabarcoding data, we additionally detected the Siberian spiny loach (Cobitis sibirica) and Ukrainian brook lamprey (Eudontomyzon mariae) and reassigned the operational taxonomic units previously assigned to Misgurnus fossilis to Cobitis sibirica. In silico analysis of metabarcoding primer efficiencies revealed considerable variability among primer pairs and among target species, which could lead to strong primer bias and potential false‐negatives in metabarcoding studies if not properly compensated for. These results highlight some of the pitfalls of eDNA‐metabarcoding as a means of monitoring fish biodiversity in large rivers, which need to be considered in order to fully unleash the full potential of these approaches for freshwater biodiversity monitoring.