基于相似日和相似时刻的变压器顶层油温预测方法

针对变压器油温受天气状况、潮流负荷等诸多因素的影响,预测精度难以提高。为解决此问题,提出了一种基于相似日和相似时刻的变压器油温预测方法,该方法是在相似日内进一步选择待预测日各时刻所对应的相似时刻,进而利用相似时刻预测变压器油温。首先采用基于气象因素的K-means聚类和时间“近大远小”原则,从历史样本中选择得到待预测日的相似日。然后在充分研究相似时刻定义描述和判断依据的基础上,基于BP神经网络和线性加权方法,给出了油温预测方法的计算步骤,并将其应用到江苏某特高压主变油温的预测工作中。最后结果表明该方法预测变压器顶层油温的精度较高，从而验证了该方法的有效性与可行性。     

高速流环境下近似连续k代表轮廓查询算法

k代表轮廓查询是从传统轮廓查询中衍生出来的一类查询.给定多维数据集合D,轮廓查询从D中找到所有不被其他对象支配的对象,将其返回给用户,便于用户结合自身偏好选择高质量对象.然而,轮廓对象规模通常较大,用户需要从大量数据中进行选择,导致选择速度和质量无法得到保证.与传统轮廓查询相比,k代表轮廓查询从所有轮廓对象中选择“代表性”最强的k个对象返回给用户,有效地解决了传统轮廓查询存在的这一问题.给定滑动窗口W和连续查询q,q监听窗口中的数据.当窗口滑动时,查询q返回窗口中,组合支配面积最大的k个对象.现有算法的核心思想是:实时监测当前窗口中的轮廓对象集合,当轮廓对象集合更新时,算法更新k代表轮廓.然而,实时监测窗口中,轮廓集合的计算代价通常较大.此外,当轮廓集合规模较大时,从中选择k代表轮廓的计算代价是同样巨大的,导致已有算法无法在高速流环境下使用.针对上述问题,提出了ρ-近似k代表轮廓查询.为了支持该查询,提出了查询处理框架PAKRS(predict-basedapproximatekrepresentativeskyline).首先,PAKRS利用高速流的特性对当前窗口进行划分,根据划分结...     

拉伸-扭转-弯曲作用下系泊缆索的应力响应

考虑6×19多股系泊缆索编制特点，研究缆索受到拉伸-扭转-弯曲共同作用下的应力响应。推导多股系泊缆索的螺旋线方程和不同层股丝接触点轴向距离，建立仿真模型中的代表性体积元素，推导该体积元素下的周期性边界条件，简化多股系泊缆索仿真模型，进一步优化应变和曲率加载方式，优化仿真模型加载形式。在有限元软件ANSYS中进行仿真分析，同时与解析模型进行对照。结果表明，所选取的代表性体积元素和周期性边界条件计算可靠，应力与曲率半径为非线性负相关关系，应变与应力为线性正相关关系，不同应变下曲率半径对应力的影响主要体现在最大应力数值以及位置，对应力-应变关系影响较小。     

Identification of Martensite Bands in Dual-Phase Steels: A Deep Learning Object Detection Approach Using Faster Region-Based-Convolutional Neural Network

Martensite banding in dual-phase steels is an important research topic in the field of materials design, since it affects the local damage properties of the material largely. Therefore, it is necessary to quantify the amount and the geometrical details of the bands in a specific microstructure, for example, for simulative approaches. A convolutional neural network is trained on manually labeled scanning electron microscopy images of DP800 steel and a subsequent effort is made to transfer these results to statistical quantities for the generation of representative volume elements (RVE). As exact geometric definitions of martensite bands in 2D are difficult, the influence of different band definitions is investigated. The result of the training shows good prediction accuracy but is strongly dependent on the chosen band definition and the underlying human bias from the labeling process. A statistical analysis using cross-validation shows that reliable results can already be achieved with only small datasets of around 50–100 training images due to the transfer learning approach. This is an important outcome as it eliminates the need to generate a large dataset which can only be obtained from time-consuming microscopy work and manual labeling of the images.     

采用"典型日"法计算日均浓度的方法探讨

“典型日”法是建设项目环境影响评价中计算日均浓度常用的方法，但由于工作经费和时间的限制，选择的典型日通常代表性较差。为此，提出了一种既减少工作量，又不失代表性的典型日确定方法，并以某建设项目环境影响评价为例进行说明。     

Geometry-based graphical methods for solar control in architecture: A digital framework

The form of a building is among the most critical design aspects concerning building energy consumption. Form-based passive design strategies, like solar control, can significantly reduce heating and cooling demands if implemented early in the design process. In this sense, there is an evident need for tools that can adequately support designers in their decisions. This paper aims to illustrate how geometry-based graphical methods (GGM) can provide effective support in the conceptual design stage. The paper introduces a novel digital framework for designing and analysing shading devices that leverages geometrical models and graphical methods. The digital implementation of GGM allows extending their applicability to three-dimensional and non-planar geometries. A comprehensive review of existing methods and tools for the design of shading devices lays the ground for the proposed digital framework, which is then demonstrated through two case studies. The results show that the diagrammatic nature of GGM facilitates a better and more direct understanding of the relationship between form and performance.     

基于代表点与K近邻的密度峰值聚类算法

密度峰值聚类(density peaks clustering, DPC)是一种基于密度的聚类算法,该算法可以直观地确定类簇数量,识别任意形状的类簇,并且自动检测、排除异常点.然而, DPC仍存在些许不足:一方面, DPC算法仅考虑全局分布,在类簇密度差距较大的数据集聚类效果较差;另一方面, DPC中点的分配策略容易导致“多米诺效应”.为此,基于代表点(representative points)与K近邻(K-nearest neighbors, KNN)提出了RKNN-DPC算法.首先,构造了K近邻密度,再引入代表点刻画样本的全局分布,提出了新的局部密度;然后,利用样本的K近邻信息,提出一种加权的K近邻分配策略以缓解“多米诺效应”;最后,在人工数据集和真实数据集上与5种聚类算法进行了对比实验,实验结果表明,所提出的RKNN-DPC可以更准确地识别类簇中心并且获得更好的聚类结果.     

黄花菜发酵液的化妆品功效评价

使用植物乳杆菌和酿酒酵母菌发酵制得了两种黄花菜发酵液，考察了样品对HaCaT细胞存活率的影响，分析了其抗炎性功效、抗氧化功效、美白功效、安全性功效。结果表明，与水提液（空白对照）相比，两种发酵液的抗氧化能力更强。在细胞水平上，两种黄花菜发酵液均表现出对UVB诱导损伤细胞的修复作用，相对于经过UVB照射后的细胞模型组，细胞Ⅰ型胶原蛋白（COL-Ⅰ）、过氧化氢酶（CAT）活力均提高。两种黄花菜发酵液的酪氨酸酶活性抑制率远高于水提液，表现出良好的美白效果。通过HaCaT细胞中炎性因子IL-6、TNF-α和COX-2表达量可以看出，样品组中炎性因子的表达量显著低于模型组，说明黄花菜发酵液的抗炎性能力大于水提液。通过鸡胚绒毛尿囊膜实验（HET-CAM）可以得出，样品作用其前后均无出血现象，表明黄花菜发酵液无刺激性。