外军地面无人装备发展现状与趋势研究

随着通信、计算机、人工智能和传感器等技术的飞速发展与广泛应用，地面无人装备日趋活跃在作战、搜救和消防等多种军民用任务场景中。对于军事领域而言，地面无人装备是陆军和其他地面作战部队迈向无人化、智能化战场的核心装备，将在未来陆战场各类作战任务中发挥不可替代的作用。通过对外军各类典型地面无人装备发展现状进行深入分析，总结出该装备领域的发展特点与趋势，对地面无人装备发展有一定的借鉴意义。     

个性化学习路径推荐综述

近年来，伴随着现代信息技术的迅猛发展，以人工智能为代表的新兴技术在教育领域得到了广泛应用，引发了学习理念和方式的深刻变革.在这种大背景下，在线学习超越了时空的限制，为学习者“随时随地”学习提供了更多的可能性，从而得到了蓬勃发展.然而，在线学习中师生时间、空间分离的特征，导致教师无法及时掌握学生的学习状态，一定程度上制约了在线学习中教学质量的提升.面对多元化的学习需求及海量学习资源，如何迅速完成学习目标、降低学习成本、合理分配学习资源等问题成为限制个人和时代发展的重大问题.然而，传统的“一刀切”的教育模式已经不能满足人们获取知识的需求了，需要一个更高效、更科学的个性化教育模式，以帮助学习者以最小的学习成本最大限度地完成学习目标.基于以上背景，如何自动高效识别学习者特征，高效地组织和分配学习资源，为每一位学习者规划个性化路径，成为面向个体的精准化教育资源匹配机制研究中亟待解决的问题.系统地综述并分析了当前个性化学习路径推荐的研究现状，并从多学科领域的角度分析了对于同一问题的不同研究思路，同时也归纳总结了当前研究中最为主流的核心推荐算法.最后，强调当前研究存在的主要不足之处.     

智能数据分区与布局研究

大数据时代,数据规模庞大,由数据进行驱动的应用分析场景日益增多.如何快速、高效地从这些海量数据中提取出用以分析决策的信息,给数据库系统带来重大挑战.同时,现代商业分析决策对分析数据的实时性要求数据库系统能够同时快速处理ACID事务和复杂的分析查询.然而,传统的数据分区粒度太粗,且不能适应动态变化的复杂分析负载;传统的数据布局单一,不能应对现代大量增加的混合事务分析应用场景.为了解决以上问题,“智能数据分区与布局”成为当前的研究热点之一,它通过数据挖掘、机器学习等技术抽取工作负载的有效特征,设计最佳的分区策略来避免扫描大量不相关的数据,指导布局结构设计以适应不同类型的工作负载.首先介绍了智能数据分区与布局的相关背景知识,然后对智能数据分区与布局技术的研究动机、发展趋势、关键技术进行详细的阐述.最后,对智能数据分区与布局技术的研究前景做出总结与展望.     

Preparation of a ternary hybrid of P-g-C3N4@PGS-Ti and its enhancement of the flame retardancy of epoxy resins

A novel ternary hybrid flame retardant named P-g-C₃N₄@PGS-Ti was prepared through step-by-step method. First, titanium dioxide was loaded on PGS to make PGS-Ti (where PGS = palygorskite), and then, PGS-Ti was decorated by phosphor-doped g-C₃N₄ (abbreviated as P-g-C₃N₄) to prepare a ternary flame retardant of P-g-C₃N₄@PGS-Ti. It showed that P-g-C₃N₄@PGS-Ti could efficiently improve the flame retardancy of epoxy resins (EP). The structure and the morphology of P-C₃N₄@PGS-Ti were characterized by Fourier transform infrared spectroscopy, X-ray diffraction, scaanning electron microscopy and hermogravimetric analysis (TGA). The flame retardancy and the burning behavior of 5 wt% P-g-C₃N₄@PGS-Ti composited EP were well investigated through TGA, limiting oxygen index (LOI), cone calorimeter test (CCT) and vertical burning test (UL-94 standard). It was found that the peak heat releasing (pk-HRR) of the EP/P-g-C₃N₄@PGS-Ti composite reduced 36% (from 1459 to 852 kW/m²) with the addition of 5 wt% of P-g-C₃N₄@PGS-Ti flame retardant to the matrix of EP. The combustion residue analysis showed that the EP/P-g-C₃N₄@PGS-Ti composite gained the most continuous and firmest char yield due to the synergistic effect of PGS, TiO₂ and the introducing of P element. The mechanism proved that the combination of gas phase and condensed phase flame-retardant processes were well coordinated to improve the fire retardancy for EP. We tested and studied the mechanical properties of EP/P-g-C₃N₄@PGS-Ti composites. Only 2.4% decreasing of flexural strength and 23.5% decreasing of impact strength in EP/P-g-C₃N₄@PGS-Ti composites compared to pure EP, respectively. But according to the test results of EP/P-g-C₃N₄@PGS-Ti composite material and the control sample in the system, EP/P-g-C₃N₄@PGS-Ti composite material had the highest flexural modulus and impact strength.     

水氯镁石制备金属镁的过程集成与能量分析

以无水氯化镁和氧化镁作为中间产物，电解和热还原为两个关键方法，集成各种相关过程，构建了从水氯镁石到金属镁的综合过程网络，其中涉及24个物种、20个化学过程和25个工艺路线；建立了最低能耗分析模型用于简单和复合过程的能量分析；利用物质的标准生成焓和多温等压摩尔热容，计算得出全部反应过程及工艺过程的能量消耗和热量移除。结果表明基于还原法的最优路径是水氯镁石用石灰法转为氢氧化镁，进而煅烧成氧化镁，再铝热还原成金属镁，该过程能耗360.15 kJ/mol，放出热量–315.46 kJ/mol；基于电解法的最优路径是石灰乳法生成氢氧化镁，再煅烧成氧化镁，通过在熔融电解质中电解生成金属镁，该过程能耗738.54 kJ/mol，放出热量–135.42 kJ/mol。无水氯化镁制备耗能高，不在最优路径中。     

Within-class multimodal classification

Multimedia Tools and Applications - In many real-world classification problems there exist multiple subclasses (or clusters) within a class; in other words, the underlying data distribution is...     

Remote Sensing Image Registration Based on Improved KAZE and BRIEF Descriptor

Machine Intelligence Research - Remote sensing image registration is still a challenging task owing to the significant influence of nonlinear differences between remote sensing images. To solve...     

Ruthenium stabilized on transition metal-on-transition metal oxide nanoparticles for naphthalene hydrogenation

The multi-metallic nanocatalysts of ruthenium nanoclusters-on-transition metal/transition metal oxide nanoparticles (TM/TMO NPs) then supported on carbon (Ru/Ni/NiO/C or Ru/Co/Co₃O₄/C) were designed and synthesized. The Ni/NiO or Co/Co₃O₄ NPs strongly stabalized the ruthenium nanoclusters by the interfacial interaction among them. These catalysts exhibited high catalytic activity and 100% selectivity to decalin for naphthalene hydrogenation due to the synergy effect of multiple catalytic sites, where naphthalene was absorbed and activated at the TMO sites (NiO or Co₃O₄), H₂ was activated at the Ru sites and it produced the activated H* species, H* was transferred to the surface of NiO or Co₃O₄ by the hydrogen spillover effect of TM (Ni or Co), reacting with the activated naphthalene and forming decalin. The nanostructures and synergetic effect of the Ru/Ni/NiO/C and Ru/Co/Co₃O₄/C catalysts were revealed by a series of techniques, such as high-resolution transmission electron microscope (HRTEM), temperature-programmed reduction (TPR), scanning transmission electron microscopy-energy dispersive X-ray spectroscopy (STEM-EDS) mapping, high-sensitivity low-energy ion scattering (HS-LEIS) and X-ray absorption spectroscopy (XAS). It is promising that the hydrogen storage can proceed at room temperature via catalyzing naphthalene hydrogenation over the Ru/Ni/NiO/C or Ru/Co/Co₃O₄/C catalyst.