Sentinel-2卫星的多光谱轻量级船舶目标检测算法

近年来深度卷积神经网络在可见光船舶检测方面取得了显著的进展，然而，大多数相关研究是通过改进大型的网络结构来提高检测性能，因此加大了对更高计算机性能的需求。此外，可见光图像难以在云、雾、海杂波、黑夜等复杂场景检测到船舶。针对以上问题，提出了一种融合红（red, R）、绿（green, G）、蓝（blue, B）和近红外（NIR）4个波段光谱信息的由粗到精细的轻量型船舶检测算法。与现有的方法中根据光谱特性利用水体检测算法提取水体区域不同之处是该算法是利用改进的水体检测算法来提取船舶候选区域。为获取更准确的候选区域，对船舶、厚云、薄云、平静海面、杂波海面5种场景中4个波段的像素值进行了统计分析，选取近红外大于阈值作为辅助判断，并以其中心点获取候选区域32×32大小的切片，并对切片进行非极大值抑制，由此获得了船舶粗检测结果。随后构建了轻量级LSGFNet网络对船舶候选区域切片进行精细识别。构建的网络融合了1×1卷积提取的波谱特征与3×3的提取几何特征，为防止光谱特征与几何特征的信息在融合时“信息不流通”，在LSGFNet网络中引入了ShuffleNet中的通道打乱机制，并减小了模型结构，与典型的轻量级网络相比具有更好的效果且模型较小。最后，利用Sentinel-2卫星多光谱10 m分辨率数据构建了512×512大小的1 120组数据进行粗检测，以及32×32大小的6 014组数据进行精细网络训练，其中候选区域粗提取的查全率为98.99%，精细识别网络精确度为96.04%，不同场景下的平均精确度为92.98%。实验表明该算法在抑制云层、海浪杂波等干扰的复杂背景下具有较高的检测效率，且训练时间短、计算机性能需求低。     

Pure tetrahedral quadruple systems with index two

An oriented tetrahedron defined on four vertices is a set of four cyclic triples with the property that any ordered pair of vertices is contained in exactly one of the cyclic triples. A tetrahedral quadruple system of order $n$ with index $\lambda$, denoted by ${\text{TQS}}_{\lambda }\left(n\right)$, is a pair $\left(X,\mathrm{?}\right)$, where $X$ is an $n$‐set and $\mathrm{?}$ is a set of oriented tetrahedra (blocks) such that every cyclic triple on $X$ is contained in exactly $\lambda$ members of $\mathrm{?}$. A ${\text{TQS}}_{\lambda }\left(n\right)$ is pure if there do not exist two blocks with the same vertex set. When $\lambda =1$, the spectrum of a pure TQS $\left(n\right)$ has been completely determined by Ji. In this paper, we show that there exists a pure ${\text{TQS}}_2\left(n\right)$ if and only if $n\equiv 1,2\left(\mathrm{mod}3\right)$ and $n\ge 7$. A corollary is that a simple ${\text{QS}}_4\left(n\right)$ also exists if and only if $n\equiv 1,2\left(\mathrm{mod}3\right)$ and $n\ge 7$.     

基于串联两阶段网络DEA模型的供应链规模收益分析

数据包络分析(DEA)是评价供应链系统(Supply chain system)间相对有效性的一种重要的工具,但是传统的DEA不考虑供应链的内部结构,对系统效率评价偏高;而本文所研究两阶段串联供应链系统,考虑把部分中间产品作为最终产品输出,增加额外中间投入的情形.基于所提出的供应链系统结构,本文建立相应的串联结构下的网络DEA模型,并针对所建立模型进行相关理论的研究,给出了串联结构下的生产可能集和规模收益情况判定方法.最后,进行数值实验,以验证我们提出的结论.     

Doxorubicin Encapsulation in Carbon Nanotubes Having Haeckelite or Stone–Wales Defects as Drug Carriers: A Molecular Dynamics Approach

Doxorubicin (DOX), a recognized anticancer drug, forms stable associations with carbon nanotubes (CNTs). CNTs when properly functionalized have the ability to anchor directly in cancerous tumors where the release of the drug occurs thanks to the tumor slightly acidic pH. Herein, we study the armchair and zigzag CNTs with Stone–Wales (SW) defects to rank their ability to encapsulate DOX by determining the DOX-CNT binding free energies using the MM/PBSA and MM/GBSA methods implemented in AMBER16. We investigate also the chiral CNTs with haeckelite defects. Each haeckelite defect consists of a pair of square and octagonal rings. The armchair and zigzag CNT with SW defects and chiral nanotubes with haeckelite defects predict DOX-CNT interactions that depend on the length of the nanotube, the number of present defects and nitrogen doping. Chiral nanotubes having two haeckelite defects reveal a clear dependence on the nitrogen content with DOX-CNT interaction forces decreasing in the order 0N > 4N > 8N. These results contribute to a further understanding of drug-nanotube interactions and to the design of new drug delivery systems based on CNTs.     

Positive ground state solutions for Schrödinger–Poisson system with critical nonlocal term in ℝ3

This paper is dedicated to studying the following Schrödinger–Poisson system $$\left\{\begin{array}{ll}-\mathrm{\Delta }u+V(x)u-K(x)\varphi |u{|}^{3}u=a(x)f(u),& x\in {ℝ}^3,\\ -\mathrm{\Delta }\varphi =K(x)|u{|}^5,& x\in {ℝ}^3.\end{array}\right.$$ Under some different assumptions on functions V(x), K(x), a(x) and f(u), by using the variational approach, we establish the existence of positive ground state solutions.     

基于深度学习的ELM实时识别研究

基于深度学习的方法，在HL-2A装置上开发出了一套边缘局域模(ELM)实时识别算法。算法使用5200次放电数据(约24.19万数据切片)进行学习，得到一个深度为22层的卷积神经网络。为衡量算法的识别能力，识别了HL-2A装置自2009年实现稳定ELMy H模放电以来所有历史数据(约26000次放电数据)，共识别出1665次H模放电，其中误识别35次，误报率为2.10%。在实际的1634次H模放电中，漏识别4次，漏识别率为0.24%。该误报率和漏报率可以满足ELM实时识别的精度要求。识别算法在实时控制环境下，对单个时间点的平均计算时间为0.46ms，可以满足实时控制的计算速度要求。     

Autonomous system for data collection: Location and mapping issues in post-disaster environment

Disaster relief requires many resources. Depending on the circumstances of each event, it is important to rapidly choose the suitable means to respond to the emergency intervention. A brief review of the conditions and means demonstrated the usefulness of an autonomous stand-alone machine for these missions. If many techniques and technologies exist, their relevant combination to achieve such a system presents several challenges. This communication tries to outline the possible achievement of an autonomous vehicle under these particular circumstances. This paper focuses on the specific working conditions and welcomes future contributions from robotics and artificial intelligence.In the necessarily limited scope of this article, the authors focus on a particularly critical aspect: location. Indeed, this machine is intended to evolve in heterogeneous and dangerous environment and without any outside contacts that could last up to several days. This blackout, due to the propagation difficulties of electromagnetic waves in the ground, induces an independence of the localisation process and makes the use of any radio navigation support system (GNSS), most of the time, impossible. The knowledge of the position of the system, both for navigation of the autonomous system (Rover) and location of targets (victims buried under debris) must be able to be estimated without contributions from external systems. Inertial classical techniques, odometer, etc., have to be adapted to these conditions during a long period without external support. These techniques also have to take into account that energy optimisation requests the use of low-power processors. Consequently, only poor computing capacity is available on-board.The article starts with a presentation of the context of a post-disaster situation as well as the main missions of Search and Rescue (SaR). It is followed by the analysis of autonomous navigation located in a post-earthquake situation. We will then discuss means to determine the attitude of the autonomous system and its position. The interest of hybridisation with external systems – whenever possible –, will be evaluated with a view to correcting deviations suffered by the system during its mission. Finally, prospects and future work are presented.     

Crosslinking network of bio‐based bis‐functional epoxides derived from trans‐limonene oxide

The thiol‐ene reaction between trans‐limonene oxide (trans‐LO) and ethane‐1,2‐dithiol in the presence of triethylborane affords a bio‐based bis‐functional epoxide (bis‐trans‐LO). The crosslinking reaction of bis‐trans‐LO with branched polyethyleneimine (BPEI; M_n = 600; BPEI₆₀₀) at a feed ratio of bis‐trans‐LO/BPEI₆₀₀ = 57/43 (wt/wt) yields the corresponding network polymer with T_d10 (10% thermal decomposition temperature) of 304.7 °C in 98% yield. In contrast, negligible amounts of network polymer are obtained by the reaction of bis‐LO (bis‐functional epoxide derived from cis and trans‐LO) and BPEI₆₀₀ regardless of the feed ratio. The mechanical strengths as measured by direct tensile tests of the network polymers derived from bis‐trans‐LO and BPEI_600,1800 (M_n = 600 and 1800) were approximately 16 and 11 times higher than that of bis‐LO and BPEI₁₈₀₀, respectively. The tensile shear strengths of the metal‐to‐metal adhesive bonds induced by bis‐trans‐LO and BPEI_600,1800 were 9.5 and 14.1 MPa, respectively. DMA revealed that the storage modulus of the network polymer derived from bis‐trans‐LO and BPEI₁₈₀₀ in the rubber region was higher than that of the material prepared from bis‐LO and BPEI₁₈₀₀, indicating higher crosslink density of the bis‐trans‐LO/BPEI₁₈₀₀ system. © 2019 Wiley Periodicals, Inc. J. Polym. Sci., Part A: Polym. Chem. 2019 , 57, 2466–2473     

腔光子-自旋波量子耦合系统中各向异性奇异点的实验研究

通过耦合三维微波腔中光子和腔内钇铁石榴石单晶小球中的自旋波量子形成腔-自旋波量子的耦合系统,并通过精确调节系统参数在该实验系统中观测到各向异性奇异点.奇异点对应于非厄米系统中一种特殊状态,在奇异点处,耦合系统的本征值和本征矢均简并,并且往往伴随着非平庸的物理性质.以往大量研究主要集中在各向同性奇异点的范畴,它的特征是在系统参数空间中沿着不同参数坐标趋近该奇异点时具有相同的函数关系.在这篇文章中,主要介绍实验上在腔光子-自旋波量子耦合系统中通过调节系统的耦合强度和腔的耗散衰减系数两条趋近奇异点的路径而实现了各向异性奇异点,具体分别对应于在趋近奇异点时,本征值的虚部的变化与耦合强度和腔的衰减系数的变化会有线性和平方根不同的行为.各向异性奇异点的实现有助于基于腔光子-自旋波量子耦合系统的量子信息处理和精密探测器件的进一步研究.