联邦学习安全与隐私保护综述

联邦学习是一种新型的分布式学习框架，它允许在多个参与者之间共享训练数据而不会泄露其数据隐私。但是这种新颖的学习机制仍然可能受到来自各种攻击者的前所未有的安全和隐私威胁。本文主要探讨联邦学习在安全和隐私方面面临的挑战。首先，本文介绍了联邦学习的基本概念和威胁模型，有助于理解其面临的攻击。其次，本文总结了由内部恶意实体发起的3种攻击类型，同时分析了联邦学习体系结构的安全漏洞和隐私漏洞。然后从差分隐私、同态密码系统和安全多方聚合等方面研究了目前最先进的防御方案。最后通过对这些解决方案的总结和比较，进一步讨论了该领域未来的发展方向。     

Modeling the effects of green alternative on heat island mitigation of a meso level town,West Bengal,India

Urban heat island (UHI) is emerged as a ubiquitous phenomenon that affects the outdoor thermal comfort. Hence urban microclimatic studies using numeric simulation software to find out adaption strategies are attracting the great attention of researchers and policymaker. Number of scientists carried out their research on this topic using different tools and technique whereas ENVI-met model is the most widely used simulation tools. The present study intended to examine the implication of green infrastructure of the town on enhanced microclimatic condition and to evaluate the best suitable mitigation strategy modelling cool city with ENVI-met software (V4). Five greening condition such as existing UHI condition (C1), 100% greening of roof (C2), 100% greening of roof and walls (C3), 50% greening of roof and walls (C4) and: Plantation at suitable area with 50% greening of roof and walls (C5) have been taken into consideration for the microclimate simulation for all three selected sites. Among these five conditions, C3 for the open mid-rise and compact low-rise, C5 for the open low-rise are identified as one of the most suitable strategies which can reduce the air temperature of peak hours by 2.6 °C, 1.33 °C and 1.87 °C respectively. These models are validated by simple linear regression between simulated and existing air temperature in case of existing UHI condition (C1) and in all the cases coefficient of determination value is high such as for open mid-rise, compact low-rise and open low-rise, it is 0.92, 0.92 and 0.75 respectively. Therefore it can be concluded that the application of those strategies can improve the urban thermal environment as well as the outdoor thermal comfort of English bazaar Town and its surroundings.     

酚醛树脂基纳米多孔材料的制备及结构调控

酚醛树脂基纳米多孔材料（Phenolic Resin-based Nanoporous Materials，PNM）是满足新一代航天飞行器轻质、高效隔热需求的新型热防护材料，传统制备方法中需使用超临界干燥技术，制备周期长、成本高。本研究通过两步法，即先合成线性酚醛树脂，再进行溶胶-凝胶的方法，实现了常压干燥PNM的制备。系统研究了固化剂含量、固化温度和固化时间对材料结构的影响和调控作用，分析了影响材料收缩率和热稳定性的因素。结果表明，PNM的微观纳米结构的变化会影响材料干燥后的收缩率，制备大颗粒、大孔径的微观结构更有利于降低材料的收缩率。而PNM的热稳定性主要受交联反应过程形成的化学结构的影响，通过优化固化剂的含量可提高PNM的热稳定性。当固化剂含量为10%，固化温度提高至150℃，固化时间延长至48 h的条件下，获得的PNM有最高的热稳定性（900℃下的残碳率为54.2%）、最发达的孔结构（比表面积为264.0 m²/g、孔容为2.67 cm³/g、平均孔径为40.0 nm）和最小的收缩率（0%）。此PNM制备方法简单、性能优异，在未来航天飞行器上有广阔的应用前景。     

多梁复合结构的动力学建模及全局解析振型分析

多梁复合结构广泛用于工程领域，其动力学特性对相关装备的工作状况有着重要影响，建立准确适用的动力学模型对后续的动力学响应参数影响规律研究有决定性作用。针对工程中常见的多梁复合结构，给出了位移场的一般表达式，利用哈密顿原理得出多梁复合结构的运动控制方程和边界条件，对其全局解析振型进行理论计算，给出了计算多梁复合结构数值振型的一般过程，基于数值分析和实验分析以L型梁和Z型梁为例利用频率相对误差、振型定性比较和MAC定量分析说明了全局解析振型的求解验证方法，利用伽辽金方法和全局解析振型给出了截断原系统得到离散系统运动方程的一般形式。提出的多梁复合结构的建模方法和全局解析振型求解验证方法，对工程中此类结构动力学响应的参数影响规律研究有重要意义。     

Mapping mean lake surface from satellite altimetry and GPS kinematic surveys

Lake water height is a key variable in water cycle and climate change studies, which is achievable using satellite altimetry constellation. A method based on data processing of altimetry from several satellites has been developed to interpolate mean lake surface (MLS) over a set of 22 big lakes distributed on the Earth. It has been applied on nadir radar altimeters in Low Resolution Mode (LRM: Jason-3, Saral/AltiKa, CryoSat-2) in Synthetic Aperture Radar (SAR) mode (Sentinel-3A), and in SAR interferometric (SARin) mode (CryoSat-2), and on laser altimetry (ICESat). Validation of the method has been performed using a set of kinematic GPS height profiles from 18 field campaigns over the lake Issykkul, by comparison of altimetry’s height at crossover points for the other lakes and using the laser altimetry on ICESat-2 mission. The precision reached ranges from 3 to 7 cm RMS (Root Mean Square) depending on the lakes. Currently, lake water level inferred from satellite altimetry is provided with respect to an ellipsoid. Ellipsoidal heights are converted into orthométric heights using geoid models interpolated along the satellite tracks. These global geoid models were inferred from geodetic satellite missions coupled with absolute and regional anomaly gravity data sets spread over the Earth. However, the spatial resolution of the current geoid models does not allow capturing short wavelength undulations that may reach decimeters in mountaineering regions or for rift lakes (Baikal, Issykkul, Malawi, Tanganika). We interpolate in this work the geoid height anomalies with three recent geoid models, the EGM2008, XGM2016 and EIGEN-6C4d, and compare them with the Mean Surface of 22 lakes calculated using satellite altimetry. Assuming that MLS mimics the local undulations of the geoid, our study shows that over a large set of lakes (in East Africa, Andean mountain and Central Asia), short wavelength undulations of the geoid in poorly sampled areas can be derived using satellite altimetry. The models used in this study present very similar geographical patterns when compared to MLS. The precision of the models largely depends on the location of the lakes and is about 18 cm, in average over the Earth. MLS can serve as a validation dataset for any future geoid model. It will also be useful for validation of the future mission SWOT (Surface Water and Ocean Topography) which will measure and map water heights over the lakes with a high horizontal resolution of 250 by 250 m.     

Substitution of satellite-based land surface temperature defective data using GSP method

Land surface temperature (LST) as an important environmental variable provides valuable information for earth environmental system modelling. Currently, LST is obtained through satellite thermal sensors at various spatial and temporal resolutions. Although spatially continuous satellite-based LST measurements are intended to overcome the shortcomings of sparse ground-based LST measurements, LST images often contain anomalous values due to the existence of clouds or sensor malfunctioning. The problem becomes more serious where the users deal with high spatial resolution characterized by low temporal resolution. This study examines the capability of a newly developed graph signal processing (GSP) method using two-dimensional single-date thermal data. For this purpose, four Landsat/TIRS datasets are analyzed. The data of five elliptical regions on thermal images are eliminated and then reconstructed through the GSP method and using the LST values of the enclosing rectangles containing the ellipsoids. The results indicate that the temperature variation determined by the GSP method generally conforms to the original image LST values. According to a correlation test conducted on the original image LST and those obtained through the GSP method, the values vary from 58% to 95%, which is an above-the-average rate (RMSE from 0.69 to 2.27). The statistical analysis of the original image LST in both the elliptical regions and the enclosing rectangles containing the ellipsoids indicates that an increase in the variance of LST data causes an increased error in the calculation of temperature by the GSP method, and vice versa. The results of the analysis of variance (ANOVA) and Duncan test indicated that an increase in the number of the non-zero spectral bins would result in increased RMSE values for all the dates and the regions. Moreover, the model errors were significant at the 0.05 level across all the image date and five elliptical study regions. Based on the results, the use of this method is recommended for the reconstruction of LST missing values, where dissimilarity of atmospheric conditions limits the use of other methods that depend on the time series data of various dates and a great deal of data calculation.     

一种针对局部结构的有限元模型修正方法

针对实际结构有限元模型（FEM）的建模误差通常仅存在于局部区域，提出了一种对局部结构单独进行模型修正的方法。首先，根据频响函数（FRF）解耦理论得到由残余结构频响函数与包含待修正参数的局部结构动刚度所重构的整体结构频响函数的拟合值，然后通过迭代优化使其与测量值的残差最小化，从而得到参数的极大似然估计。在此基础上，将残差关于参数的灵敏度以局部结构动力学矩阵表示，建立了模型修正的基本方程，利用整体结构的测试数据即可直接对分离出来的局部结构进行模型修正。最后，对喷气式飞机和三角机翼飞机分别进行了数值模拟和实验研究，验证了所提方法的可行性和有效性。结果表明，所提方法可以成功地用于仅局部区域含有建模误差的实际结构有限元模型的修正，修正后的有限元模型的动态特性与实际结构有较好的一致性。     

Manifold structure preservative for hyperspectral target detection

A nonparametric method termed as manifold structure preservative (MSP) is proposed in this paper for hyperspectral target detection. MSP transforms the feature space of data to maximize the separation between target and background signals. Moreover, it minimizes the reconstruction error of targets and preserves the topological structure of data in the projected feature space. MSP does not need to consider any distribution for target and background data. So, it can achieve accurate results in real scenarios due to avoiding unreliable assumptions. The proposed MSP detector is compared to several popular detectors and the experiments on a synthetic data and two real hyperspectral images indicate the superior ability of it in target detection.     

多种工质下螺旋波等离子体源波场结构数值模拟

针对螺旋波等离子体放电机理，开展了多种工质条件下的螺旋波放电等离子体内波场结构数值模拟研究。计算发现:氦气等离子体的Er分量在径向边界处的峰值更为突出，有利于等离子体在径向的输运，波电场径向分量决定了电流密度径向分量在内部的表现。在0.266 Pa和1.064 Pa两种气体压强条件下，通过波场结构验证了气压对于波阻尼影响的结论。波场结构是螺旋波在等离子体内传播以及能量沉积的微观体现，研究螺旋波波场结构是揭示其高电离效率的重要途径。初步探索了功率耦合机制，为实验系统优化及实验方案设定奠定理论基础。     

一种基于事件序列的故障树最小割集算法

在系统分析故障树最小割集各种求解算法的基础上，提出了一种故障树事件序列的数据结构，并针对该数据结构，设计了一种高效的故障树最小割集求解算法，并对其替换子算法、复制子算法和消减子算法进行了形式化描述。算法可有效降低最小割集生成中的“组合爆炸”问题，实验结果表明算法具有较好的计算性能。