隧道安全预警机器人自主导航方法研究

针对隧道施工环境,设计了一款隧道安全预警机器人,并重点研究其导航系统实现;机器人搭载检测仪器与分析主机,集数据采集、分析、预警于一体,实现隧道巡检的自动化和智能化;分别利用Cartographer算法和自适应蒙特卡洛算法(AMCL)融合雷达、里程计、IMU的数据,实现环境地图的创建与机器人实时定位;采用A * 算法和DWA算法实现机器人的全局和局部路径规划,并结合ROS操作系统对其进行实验验证;实验结果表明:在导航定位时,机器人位置误差和航向误差分别小于20 cm和5.5.,能满足实际工程需要,完成自动巡检任务.     

Stabilizing effect of the cyclodextrins additive to spray-dried particles of curcumin/polyvinylpyrrolidone on the supersaturated state of curcumin

Although curcumin is considered to have various therapeutic effects, its use as a functional food or supplement is restricted owing to its low water solubility and bioavailability. To increase the solubility of curcumin in water, the use of polyvinylpyrrolidone (PVP) and vinylpyrrolidone-vinyl acetate copolymers with a pyrrolidone skeleton was noted to be promising. In particular, the bi-component formulations of curcumin/PVP prepared through spray drying exhibited an amorphous state in powder X-ray diffraction observations and temporally increased the apparent solubility of curcumin to over 5000 times that of untreated curcumin; nevertheless, after 24 h, the solubility decreased owing to the unstable supersaturated state of curcumin. The addition of α-cyclodextrin (α-CyD) in the bi-component curcumin/PVP formulation helped maintain the supersaturated state of curcumin, whereas the addition of β- and γ-CyD led to the collapse of the supersaturated state. The addition of α-CyD can likely help inhibit the nucleation and crystal growth of curcumin, through the interaction among the solubilized units of curcumin/PVP and α-CyD.     

移动机器人同步定位与建图技术综述

同步定位与建图技术(SLAM)一直是移动机器人领域比较热门的研究方向,它可以给机器人提供强大的环境感知能力;传统的依靠外部位置参考来定位的方法如果无法获得时,移动机器人需要即时定位自身位置来构建增量式地图,因此SLAM技术也就应运而生;对激光SLAM和视觉SLAM的研究现状及最新标志性成果进行了介绍,重点对以相机与激光雷达融合、相机与IMU融合、激光雷达与IMU融合为代表的多传感器融合SLAM技术展开讨论、系统地梳理了几种融合方式的优势与不足,同时介绍了该领域的研究热点语义SLAM,最后讨论了SLAM技术在该领域未来的发展方向以及存在的挑战。     

Multi-source information fusion for smart health with artificial intelligence

The rapid developments in Artificial Intelligence present an opportunity for the research community to provide and advance Smart Health for the well-being of our society. By considering the availability of multi-source information and heterogeneous data in the era of Big Data, this Special Issue explores the theories, methodologies and possible breakthroughs that have designed and adopted information fusion for Smart Health powered by recent Artificial Intelligence advances. Specifically, this Special Issue focuses on three questions; How to achieve and realize human-level intelligence in Smart Health, How to achieve and benefit Smart Health from a multi-disciplinary balance, and How to utilize the power of Big Data for Smart Health. The Special Issue is a great success, with a small number of quality studies carefully selected from an overwhelming amount of contributions.     

Alloy design via additive manufacturing: Advantages,challenges, applications and perspectives

Additive manufacturing (AM) has rapidly changed both large- and small-scale production environments across many industries. By re-envisioning parts from the ground up, not limited to the challenges presented by traditional manufacturing techniques, researchers and engineers have developed new design strategies to solve large-scale materials and design problems worldwide. This is particularly true in the world of alloy design, where new metallic materials have historically been developed through tedious processes and procedures based primarily on casting methodologies. With the onset of directed energy deposition (DED) and powder bed fusion (PBF)-based AM, new alloys can be innovated and evaluated rapidly at a lower cost and considerably shorter lead time than has ever been achieved. This article details the advantages, challenges, applications, and perspectives of alloy design using primarily laser-based AM. It is envisioned that researchers in industry and academia can utilize this work to design new alloys leveraging metallic AM processes for various current and future applications.     

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.     

基于节点1-邻居图相似性的社会网络匿名技术

利用传统的k匿名技术在社会网络中进行隐私保护时会存在聚类准则单一、图中数据信息利用不足等问题. 针对该问题, 提出了一种利用Kullback-Leibler (KL)散度衡量节点1-邻居图相似性的匿名技术(anonymization techniques for measuring the similarity of node 1-neighbor graph based on Kullback-Leibler divergence, SNKL). 根据节点1-邻居图分布的相似性对原始图节点集进行划分, 按照划分好的类进行图修改, 使修改后的图满足k匿名, 完成图的匿名发布. 实验结果表明, SNKL方法与HIGA方法相比在聚类系数上的改变量平均降低了17.3%, 同时生成的匿名图与原始图重要性节点重合度保持在95%以上. 所提方法在有效保证隐私的基础上, 可以显著的降低对原始图结构信息的改变.