倾斜入射下的单截面平面度绝对检验

针对大尺寸光学平面的直线度的纳米级测量精度需求，提出了倾斜入射下单截面平面度绝对检验方法，实现了对超过相移干涉仪口径的长平晶绝对检验。利用棱镜转向实现倾斜入射角度的精密预标定，棱镜标定角度的精度高于圆光栅和图像分析等方法，可提高测量不确定度到0.0042 μm。对比了常规三面互检绝对检验结果与本方法的差异，在相同尺寸下，直线度误差仅为1.2 nm。在确认标定反射镜位置后，整个倾斜入射的干涉图调整过程将被完全集中到待测长平晶的工作面上，不需要再对反射平晶进行操作。调整长平晶时各个维度的操作互不干涉，可快速简便地得到测量结果。     

Quality assessment for color correction-based stitched images via bi-directional matching

With the deepening of social information, the panoramic image has drawn a significant interest of viewers and researchers as it can provide a very wide field of view (FoV). Since panoramic images are usually obtained by capturing images with the overlapping regions and then stitching them together, image stitching plays an important role in generating panoramic images. In order to effectively evaluate the quality of stitched images, a novel quality assessment method based on bi-directional matching is proposed for stitched images. Specifically, dense correspondences between the testing and benchmark stitched images are first established by bi-directional SIFT-flow matching. Then, color-aware, geometric-aware and structure-aware features are respectively extracted and fused via support vector regression (SVR) to obtain the final quality score. Experiments on our newly constructed database and ISIQA database demonstrate that the proposed method can achieve comparable performance compared with the conventional blind quality metrics and the quality metrics specially designed for stitched images.     

2种侧围加强内框总成检具结构对比分析

概述了侧围加强内框总成与周边零件的匹配关系,从检具操作的人机工程出发,描述了侧围加强内框总成检具的2种定位方式和结构设计,对比了2种定位方式对于检测结果的影响,总结了其优点和弊端,结果表明侧围加强内框总成检具采用车内方向朝下的定位方式效果更好。     

Synthesis of Hollow Three-Dimensional Channels LiNi_(0.5)Mn_(1.5)O_4 Microsphere by PEO Soft Template Assisted with Solvothermal Method

A appropriate size with three-dimension(3 D) channels for lithium diffusion plays an important role in constructing highperforming LiNi_(0.5)Mn_(1.5)O_4(LNMO) cathode materials, as it can not only reduce the transport path of lithium ions and electrons, but also reduce the side effects and withstand the structural strain in the process of repetitive Li~+ intercalation/deintercalation. In this work, an e fficient method for designing the hollow LNMO microsphere with 3 D channels structure by using polyethylene oxide(PEO) as soft template agent assisted solvothermal method is proposed. Experimental results indicate that PEO can make the reagents mingle evenly and nucleate slowly in the solvothermal process, thus obtaining a homogeneous distribution of carbonate precursors. In the final LNMO products, the hollow 3 D channels structure obtained by the decomposition of PEO and carbonate precursor in the calcination can provide abundant electroactive zones and electron/ion transport paths during the charge/discharge process, which benefits to improve the cycling performance and rate capability. The LNMO prepared by adding 1 g PEO possesses the most outstanding electrochemical performance, which presented an excellent discharge capacity of 143.1 mAh g~(-1) at 0.1 C and with a capacity retention of 92.2% after 100 cycles at 1 C. The superior performance attributed to the 3 D channels structure of hollow microspheres, which provide uninterrupted conductive systems and therefore achieve the stable transfer for electron/ion.     

Labeling and Characterization of Phenol-Containing Glycopeptides Using Chemoselective Probes with Isotope Tags

Secondary metabolites are structurally diverse natural products (NPs) and have been widely used for medical applications. Developing new tools to enrich NPs can be a promising solution to isolate novel NPs from the native and complex samples. Here, we developed native and deuterated chemoselective labeling probes to target phenol-containing glycopeptides by the ene-type labeling used in proteomic research. The clickable azido-linker was included for further biotin functionalization to facilitate the enrichment of labeled substrates. Afterward, our chemoselective method, in conjunction with LC-MS and MSn analysis, was demonstrated in bacterial cultures. A vancomycin-related phenol-containing glycopeptide was labeled and characterized by our labeling strategy, showing its potential in glycopeptide discovery in complex environments.     

Multilayer Si@SiOx@void@C anode materials synthesized via simultaneously carbonization and redox for Li-ion batteries

In the development of high-performance lithium-ion batteries (LIBs), the composition and structure of electrode materials are of critical importance. Silicon has a theoretical specific capacity 10 times that of graphite, nonetheless, its application as an anode material confronts challenge as it undergoes huge volume change and pulverization amidst the alloying and dealloying processes. Herein, a novel method to prepare a multilayer Si-based anode was proposed. Three layers, SiO₂, nickel and triethylene glycol (TEG), were coated successively on Si nanoparticles, which served respectively as the sources of SiO_x, sacrificial templates and carbon. Nickel can not only serve as a hollow template, but also play a catalytic role, which makes carbonization and redox reactions occur synchronously under a mild condition. Amid the carbonization process of TEG at 450 °C, several-nm-thick SiO₂ layer can react with the as-derived carbon to form a silicon suboxides (SiO_x (0 < x < 2)) intermedium layer. After removing the nickel template, a micro-nano scaled Si@SiO_x@void@C with conformal multilayer-structure can be obtained. The BET specific surface area and pore volume of powders were increased dramatically because of the derivation of abundant voids, which can not only buffer the swelling effect of silicon, but also provide richer ionic conductivity. The as-assembled half-cell with Si@SiOx@void@C as the anode material possesses high capacity (～1000 mAh g^?1 at 3 A g^?1), long cycle life (300 cycles with 77% capacity retention) and good rate performance (558 mAh g^?1 at 5 A g^?1).     

结合预训练模型和语言知识库的文本匹配方法

针对文本匹配任务,该文提出一种大规模预训练模型融合外部语言知识库的方法。该方法在大规模预训练模型的基础上,通过生成基于WordNet的同义—反义词汇知识学习任务和词组—搭配知识学习任务引入外部语言学知识。进而,与MT-DNN多任务学习模型进行联合训练,以进一步提高模型性能。最后利用文本匹配标注数据进行微调。在MRPC和QQP两个公开数据集的实验结果显示,该方法可以在大规模预训练模型和微调的框架基础上,通过引入外部语言知识进行联合训练有效提升文本匹配性能。     

面向智能客服系统的情感分析技术

该文以阿里小蜜为例,对智能客服系统中的情感分析技术进行比较全面的介绍,包括情感分析算法模型的原理及其在智能客服系统的多个应用场景中的实际落地使用方式和效果分析。智能客服在解决客户高频业务问题的同时,也需要给客户提供多维度的、具有类人能力的助理、导购、语聊和娱乐等服务,提高客户对智能客服机器人的整体满意度。在此过程中,情感分析技术在机器人类人能力建设中起到了至关重要的作用。该文围绕智能客服系统中人机结合的服务形式,从六个维度总结和介绍了情感分析技术在智能客服系统中的应用场景,包括用户情感检测、用户情感安抚、情感生成式语聊、客服服务质检、会话满意度预估和智能人工入口。     

储能水管-冲击式水轮机发电系统设计及发电效益分析

为充分利用高层建筑物内部的水能包括生活污水及高层建筑物顶楼雨水,可将其转化为电能。由此设计的储能水管-冲击式水轮机装置即通过冲击式水轮机和发电机将高层建筑物内部储存的生活污水及雨水的势能转变为可供使用的电能,并通过建立单户用水模拟模型,设置储能水管对30、100层建筑进行发电模拟。模拟结果表明,储能水管-冲击式水轮机发电具有很好的发电效益和节能减排社会效益,获得了日发电过程中冲击式水轮机的水头变换和启停特性,为该系统中冲击式水轮机的个性设计提供了依据。