Presentation attack detection based on two-stream vision transformers with self-attention fusion

Aiming at the performance degradation of the existing presentation attack detection methods due to the illumination variation, a two-stream vision transformers framework (TSViT) based on transfer learning in two complementary spaces is proposed in this paper. The face images of RGB color space and multi-scale retinex with color restoration (MSRCR) space are fed to TSViT to learn the distinguishing features of presentation attack detection. To effectively fuse features from two sources (RGB color space images and MSRCR images), a feature fusion method based on self-attention is built, which can effectively capture the complementarity of two features. Experiments and analysis on Oulu-NPU, CASIA-MFSD, and Replay-Attack databases show that it outperforms most existing methods in intra-database testing and achieves good generalization performance in cross-database testing.     

一种用于电气设备状态监测的新式FBG传感器北大核心CSCD

为了监测绕组变压器的静态应力场和发生短路等故障时的动态应力变化,设计了一种用于电气设备状态监测的新式FBG传感器。该传感器由聚醚醚酮材料封装的FBG构成,通过内部圆锥形空腔结构实现将轴向应力集中于FBG敏感位置。通过仿真对不同压力强度下传感器结构的应力场部分及形变趋势进行了计算与分析,论证了设计的合理性。实验分别对静态载荷和动态冲击进行测试,结果显示,在静态压载测试中,当100 N相似文献   

Korean automatic spacing using pretrained transformer encoder and analysis

Automatic spacing in Korean is used to correct spacing units in a given input sentence. The demand for automatic spacing has been increasing owing to frequent incorrect spacing in recent media, such as the Internet and mobile networks. Therefore, herein, we propose a transformer encoder that reads a sentence bidirectionally and can be pretrained using an out-of-task corpus. Notably, our model exhibited the highest character accuracy (98.42%) among the existing automatic spacing models for Korean. We experimentally validated the effectiveness of bidirectional encoding and pretraining for automatic spacing in Korean. Moreover, we conclude that pretraining is more important than fine-tuning and data size.     

基于变压器最优漏磁信息的绕组变形分类方法

为了判别电力变压器绕组变形状况,提出一种基于最优漏磁信息及多层聚类编码和最优权重译码的ECOC分类器对变压器绕组变形分类的方法,首先利用多层聚类算法构建最优ECOC编码矩阵,然后通过最优权重译码算法对分类器输出进行译码得出变压器绕组具体变形类别,建立变压器绕组二维漏磁场有限元模型,计算得出绕组可能出现的变形形式和绕组区域变形前后的磁感应强度值数据,从而得出磁感应强度测量点;最后利用所建立模型得出的绕组变形数据进行仿真判别。结果表明,所建立的分类器在绕组变形判别时具有较高的准确性,可用于变压器绕组变形类型的检测。     

A new design strategy for DC/DC LLC resonant converter: Concept,modeling, and fabrication

In this paper, a new design procedure for LLC converter has been introduced. In fact, this method is a computer-based design algorithm based on a numerical technique. In the process of designing, the value of the resonant element is obtained by solving the LLC converter fundamental equation. This converter will be controlled by using state feedback, such as output voltage variable. As a matter of fact, in a control system, the change of output voltage (because of load variation) will affect the switching frequency, so the output voltage will be tuned. In the designing process, the fundamental equations of LLC converter are obtained, and the value of the resonant elements is calculated. Also, a comparison analysis is carried out between the proposed and typical methods. The simulation is done to investigate the validity of the proposed method. Moreover, a prototype is manufactured, and the experimental test is done to evaluate its applicability.     

A fuel-cell/battery hybrid DC backup power system via a new high step-up three port converter

In this paper, a fuel-cell (FC)/battery hybrid direct-current (DC) backup power system is proposed for high step-up applications. This system is composed of a newly developed non-isolated three-port converter, which achieves a high voltage gain by taking the advantage of a quasi Z-source network and an energy transfer capacitor. After analyzing the converter, a comprehensive comparison study and a design procedure are provided. Moreover, the controllers regulating the source power levels while smoothing the FC power profile according to the proposed energy management strategy (EMS) are designed based on the developed small-signal model of the proposed converter. Both hardware and controller design procedures are validated through the PSIM model of the whole system. As a result, it is shown that the proposed system can effectively couple FC and battery while transferring their energies to a high voltage DC bus according to the offered EMS.     

不同芳烃种类对变压器油性能的影响

考察了不同芳烃种类对变压器油电气性能、抗氧化性能、溶解性能、抗析气性能和热物性的影响。结果表明：重芳烃、中芳烃、轻芳烃对变压器油电气性能影响由大到小的顺序为：重芳烃>中芳烃>轻芳烃；不同芳烃对基础油抗氧化性能和热物性影响不大，但中芳烃和重芳烃对抗氧剂感受性很差，而轻芳烃影响较小；溶解性能主要与芳烃含量有关，受芳烃种类影响较小，芳烃含量越高，溶解性能越好；芳烃具有良好的抗析气性，且中芳烃对基础油抗析气性的改善明显优于轻芳烃和重芳烃；在保证优异的电气性能和抗氧化性能前提下，变压器油中含适量的芳烃可改善变压器油的溶解性能和抗析气性能。     

简析变压器油的生产工艺及质量

本文简要介绍了变压器油的运用价值,同时分析如何实现对变压器油生产工艺的优化,并从物理性能以及化学性能两个方面分析生产工艺改良后变压器油整体质量,以期为变压器油生产工作提供一定的参考与帮助。     

FR3绝缘油配电变压器的过载能力评估

为提高配电变压器过载性能,基于传热学基本理论与《油浸式电力变压器负载导则》中的油浸式电力变压器过载运行的热点温度计算模型,提出适用于FR3绝缘油配电变压器的热点温度模型,根据理论计算结果比较了传统矿物油配电变压器与FR3绝缘油配电变压器的性能差异。结果表明,相同结构下,FR3绝缘油配电变压器能在1.27倍额定负载下长期运行,短期过载运行的安全时间远高于传统矿物油配电变压器,表现出更强的过载能力,有广阔的应用前景。