Current measurement method based on magneto-optic rotation effect

Aiming at the approximate measurement of magnetic rotation angle in optical current sensor based on light intensity detection mode, this paper proposes a current measurement method based on triangular constant transformation to reconstruct magnetic rotation angle, so as to avoid the large current measurement error caused by the approximate measurement of the magnetic rotation angle. By extracting the direct current (DC) component and the alternating current (AC) component of the light intensity signal detected by the photoelectric detector (PD), the sine signal containing the magnetic rotation angle is directly obtained by dividing the two components, and then the triangular identity transformation method is used to linearly demodulate the magnetic rotation angle and reconstruct the current waveform. The experimental results show that the relative error of current measurement does not exceed 1.40% in the current range of 0.05—0.50 A, which is less than the approximate linear measurement (ALM) method, and the magnetic rotation angle and the current have a good linear relationship.     

利用运动向量差值改善深度学习视频隐写分析

目的 针对现有深度学习视频隐写分析网络准确率不够高的问题,本文从视频压缩编码的原理出发,发掘嵌密编码参数与其他参数之间的关系,通过拓展检测空间,构造新的检测通道,改善现有深度学习视频隐写分析网络的检测性能。方法 以H. 265/HEVC(high efficiency video coding)压缩视频为例,首先通过分析运动向量的嵌密修改对运动向量差值的影响,指出可将运动向量差值作为新增的采样对象(或称检测对象);接着,提出一个构造运动向量差值检测矩阵的方法,解决了空域上采样样本稀疏、时域上样本空间位置无法对齐的问题;最后,将运动向量差值矩阵直接用于改善现有的VSRNet(video steganalysis residual network)、SCA-VSRNet(selection-channel-aware VSRNet)以及Q-VSRNet(quantitative VSRNet)等3个H. 265/HEVC深度学习视频隐写分析网络,分别得到IVSRNet(improved VSRNet)、SCA-IVSRNet(selection-channel-aware improved...     

注意力增强和目标模型更新的红外目标跟踪算法

目的 多数以深度学习为基础的红外目标跟踪方法在对比度弱、噪声多的红外场景下,缺少对目标细节信息的利用,而且当跟踪场景中有相似目标且背景杂乱时,大部分跟踪器无法对跟踪的目标进行有效的更新,导致长期跟踪时鲁棒性较差。为解决这些问题,提出一种基于注意力和目标模型自适应更新的红外目标跟踪算法。方法 首先以无锚框算法为基础,加入针对红外跟踪场景设计的快速注意力增强模块以并行处理红外图像,在不损失原信息的前提下提高红外目标与背景的差异性并增强目标的细节信息,然后将提取的特征融合到主干网络的中间层,最后利用目标模型自适应更新网络,学习红外目标的特征变化趋势,同时对目标的中高层特征进行动态更新。结果 本文方法在 4 个红外目标跟踪评估基准上与其他先进算法进行了比较,在 LSOTB-TIR(large-scale thermalinfrared object tracking benchmark)数据集上的精度为 79.0%,归一化精度为 71.5%,成功率为 66.2%,较第 2 名在精度和成功率上分别高出 4.0%和 4.6%;在 PTB-TIR(thermal infrared pedestrian tracking benchmark)数据集上的精度为85.1%,成功率为 66.9%,较第 2 名分别高出 1.3% 和 3.6%;在 VOT-TIR2015(thermal infrared visual object tracking)和VOT-TIR2017 数据集上的期望平均重叠与精确度分别为 0.344、0.73 和 0.276、0.71,本文算法在前 3 个数据集的测评结果均达到最优。同时,在 LSOTB-TIR 数据集上的消融实验结果显示,本文方法对基线跟踪器有着明显的增益作用。结论 本文算法提高了对红外目标特征的捕捉能力,解决了红外目标跟踪易受干扰的问题,能够提升红外目标长期跟踪的精度和成功率。     

Regulation of Ionic Distribution and Desolvation Activation Energy Enabled by In Situ Zinc Phosphate Protective Layer toward Highly Reversible Zinc Metal Anodes

Despite the merits of high specific capacity, low cost, and high safety, the practical application of aqueous Zn metal batteries (AZMBs) is plagued by the dendritic growth and corrosion reaction of Zn metal anodes. To solve these issues, a Zn₃(PO₄)₂·4H₂O protective layer is in-situ constructed on Zn foil (Zn@ZnPO) by a simple hydrothermal method, avoiding the traditional slurry-casting process. The insulating and conformable ZnPO layer improves the wettability of Zn@ZnPO and aqueous electrolyte via decreasing the contact angle to 11.7^o. Compared with bare Zn, the Zn@ZnPO possesses a lower desolvation activation energy of 35.25 kJ mol^-1, indicating that the ZnPO fasters the desolvation of hydrated Zn²⁺ ions and thereby ameliorates their transport dynamics. Micro-morphology and structural characterization show that there are no dendrites forming on the post-cycling Zn@ZnPO anodes, and the interfacial ZnPO layer remains almost identical before and after cycles. It can be explained that the electrochemically stable ZnPO layer acts as an ionic modulator to enable the homogeneous distribution of Zn²⁺ ions, inhibiting the growth of Zn dendrites. Benefiting from these advantages, the Zn@ZnPO based symmetric and full cells deliver highly reversible Zn plating/stripping behavior and long cycling lifespans.     

物理域中针对人脸识别系统的对抗样本攻击方法

对抗样本攻击揭示了人脸识别系统可能存在不安全性和被攻击的方式。现有针对人脸识别系统的对抗样本攻击大多在数字域进行,然而从最近文献检索的结果来看,越来越多的研究开始关注如何能把带有对抗扰动的实物添加到人脸及其周边区域上,如眼镜、贴纸、帽子等,以实现物理域的对抗攻击。这类新型的对抗样本攻击能够轻易突破市面上现有绝大部分人脸活体检测方法的拦截,直接影响人脸识别系统的结果。尽管已有不少文献提出数字域的对抗攻击方法,但在物理域中复现对抗样本的生成并不容易且成本高昂。本文提出一种可从数字域方便地推广到物理域的对抗样本生成方法,通过在原始人脸样本中添加特定形状的对抗扰动来攻击人脸识别系统,达到误导或扮演攻击的目的。主要贡献包括:利用人脸关键点根据脸型构建特定形状掩膜来生成对抗扰动;设计对抗损失函数,通过训练生成器实现在数字域的对抗样本生成;设计打印分数损失函数,减小打印色差,在物理域复现对抗样本的生成,并通过模拟眼镜佩戴、真实场景光照变化等方式增强样本,改善质量。实验结果表明,所生成的对抗样本不仅能在数字域以高成功率攻破典型人脸识别系统VGGFace10,且可方便、大量地在物理域复现。本文方法揭示了人脸识别系统的潜在安全风险,为设计人脸识别系统的防御体系提供了很好的帮助。     

Stubborn state estimation for complex-valued neural networks with mixed time delays: the discrete time case

Neural Computing and Applications - In this paper, the state estimation problem is investigated for a class of discrete-time complex-valued neural networks (CVNNs) with mixed time delays. We...     

技术热点、前沿识别支持的2035技术清单调整方法——以机器人技术为例

本文针对技术预见中关注度较高的技术热点、前沿识别和趋势分析问题,综合以往研究提出的成体系的和较为完善的分析方法,并以机器人技术为例,分析了机器人领域的技术热点和技术前沿以及发展趋势,为中国工程科技2035发展战略研究中技术清单的提出与技术产业布局提供了重要的数据支持。     

Smart Utilization of Carbon Dots in Semiconductor Photocatalysis

Efficient capture of solar energy will be critical to meeting the energy needs of the future. Semiconductor photocatalysis is expected to make an important contribution in this regard, delivering both energy carriers (especially H₂) and valuable chemical feedstocks under direct sunlight. Over the past few years, carbon dots (CDs) have emerged as a promising new class of metal‐free photocatalyst, displaying semiconductor‐like photoelectric properties and showing excellent performance in a wide variety of photoelectrochemical and photocatalytic applications owing to their ease of synthesis, unique structure, adjustable composition, ease of surface functionalization, outstanding electron‐transfer efficiency and tunable light‐harvesting range (from deep UV to the near‐infrared). Here, recent advances in the rational design of CDs‐based photocatalysts are highlighted and their applications in photocatalytic environmental remediation, water splitting into hydrogen, CO₂ reduction, and organic synthesis are discussed.     

基于智能传感器MPXY8300的TPMS设计

介绍了当前轮胎压力监测系统的设计方法,提出了基于MPXY8300智能传感器的轮胎压力监测模块和基于MC33596接收机的接收模块的设计方案,并给出了硬件电路以及软件程序流程。该系统可随时测定每个轮胎内部的实际温度和瞬压,确定故障轮胎并实时发出警示,有效避免爆胎事故的发生。