基于随机森林特征变量优化的湿地植物分类与密度反演

以长江口滨海湿地为研究区域,采用随机森林算法对滨海湿地植被进行分类。在提取Landsat?8 OLI影像植被指数和水体指数的基础上,提出利用植被指数季节差值对模型进行特征变量优化,分析了长江口滨海湿地植物群落分布的空间特征。以所占面积最大的互花米草（入侵物种）为例,采用多元线性回归模型结合实地测量数据,估算了秋季的互花米草植物密度的空间特征。提出的多时相遥感数据结合随机森林特征变量优化方法,可以较为便捷地提取长江口湿地3种优势物种的空间分布特征,与最大似然法相比,分类精度有较大提高,总体分类精度由78.35%提高至87.55%,Kappa系数由0.72提高至0.84。该方法适用于存在“异物同谱”问题的湿地植物群落研究。     

2, 6-二亚苄基环己酮液晶化合物的合成及性能研究

以苯甲醛、环己酮为原料, 四丁基溴化铵为相转移催化剂, 在氢氧化钠溶液中合成了具有液晶性质的化合物2,6-二亚苄基环己酮. 考察了反应时间、温度、碱浓度对产率的影响. 产品经红外光谱、核磁共振氢谱以及元素分析予以表征, 用差示扫描量热(DSC)、偏光显微镜和流变仪研究了其液晶行为, 证明为近晶相热致液晶. 对化合物晶体进行了X射线单晶衍射测定, 证实属于单斜晶系, 空间群P2(1)/c, a = 9.586(1) Å, b = 18.391(2) Å, c = 9.433(1) Å, α = 90°, β = 115.816(2)°, γ = 90°, Dc = 1.217 g·cm^-3, V = 1496.9(3) Å ^-3, Z = 4.     

新型极化冷冻油添加剂持续改善空调系统性能实验研究

对一台热泵型空调器添加一种新型极化冷冻油添加剂(PROA),进行了为期40天的标准工况下的性能测试跟踪.实验结果表明,与添加PROA之前相比,压缩机的吸气、排气温度等一些参数明显下降;PROA加入后的运行期间,压缩机吸气、排气温度变化不大,波动仅1.5℃左右,吸气、排气压力和室内、外出风温度也都变化不大,且对压缩机等核心部件还未见有何负面作用.说明了添加PROA对空调系统的影响基本稳定.     

铁磁/有机系统的自旋扩散漂移性质研究

基于自旋扩散漂移方程和欧姆定律,理论研究了电场对铁磁/有机半导体界面的电流自旋极化性质的影响.考虑到有机半导体内特殊的载流子以及电场对其自旋扩散长度的影响,计算了界面处的电流自旋极化率.结果表明,高电场可以使界面处的电流自旋极化率得到有效提高.同时还进一步研究了电场下有机半导体中极化子比率、自旋相关界面电阻等因素对电流自旋极化的影响.     

基于三次模型的线搜索方法及其收敛性

针对无约束优化问题,给出一种基于三次模型的线搜索型算法,并在适当的条件下证明算法的全局收敛性.该算法以对称矩阵代替原三次模型中的Hessian阵,并且不需要保持正定和Dennis-Moré条件,它与一般线搜索法不同,在每次迭代中步长可以在下降方向上由显性公式直接确定,从而可以减少搜索计算.     

凹坑结构对麻面管流动与传热特性的影响分析

采用三维数值模拟方法,分析不同进口油速(0. 6～1. 8 m/s)、凹坑高径比(0. 45、0. 474、0. 5、0. 529、0. 563、0. 6)对麻面管流动与传热特性的影响。结果表明:高径比越小,凹坑内旋涡区越扁平;不同高径比凹坑内旋涡位置不同,同时对麻面管性能的影响也不同;相同进口油速下,麻面管的努塞尔数均高于光管,阻力系数均低于光管;相同进口油速下,随高径比增加,努塞尔数先增加后减小,阻力系数先减小后增加;麻面管在凹坑高径比为0. 5时综合传热效果最好,此时为最优凹坑结构。     

Lightweight Privacy-Aware Yet Accountable Secure Scheme for SM-SGCC Communications in Smart Grid

Smart grid is envisioned as a critical application of cyber-physical systems and of the internet of things.In the smart grid,smart meters equipped with wireless sensors can upload meter readings (data) to smart grid control and schedule centers via the advanced metering infrastructure to improve power delivery efficiency.However,data gathered in short intervals,such as 15minutes,will expose customers’ detailed daily activities (for example,when they get up and when they use oven) using nonintrusive appliance load monitoring.Thus,data must be hidden to protect customers’ privacy.However,data accountability is still required for emergency responses or to trace back suspected intrusions,even though the data is anonymous.In addition to desired security requirements,this imposes two extra tasks:Sensors in smart meters usually have resource constraints;thus,the desired security protocols have to remain lightweight in terms of computation and storage cost.Furthermore,scalability and flexibility are required since there exist vast meters.This paper presents a lightweight Privacy-aware yet Accountable Secure Scheme called PASS which guarantees privacy-aware accountability yet tackles the above challenges in the smart grid.A formal security analysis justifies that PASS can attain the security goals,while a performance analysis verifies that PASS requires few computations,and is scalable and flexible.     

Degree of allogeneic histoincompatibility assayed by the hemolytic plaque technique

Zusammenfassung Es wurden in vitro die Wechselwirkungen von Milzzellen verschiedener allogener Mausstämme untersucht. Als Indikator für die cytotoxische Wirkung der lymphoiden Zellen wurde die Fähigkeit dieser Zellen, in vitro Antikörper zu bilden, herangezogen (Methode vonMishell undDutton). Es hat sich dabei gezeigt, dass bei ausgeprägter gegenseitiger Histoinkompatibilität die Zahl der Plaque formenden Zellen erniedrigt ist. Bei geringer Histoinkompatibilität war die Zahl der Plaque formenden Zellen unverändert oder leicht erhöht.

---

---

On leave of absence from National Institute of Radiological Sciences, Chiba (Japan).     

Suppression of humoral immune response in mice by administration of high molecular levan

Summary High molecular levan, a polyfructoside, has a dose-dependent inhibitory effect on the primary immune response to sheep red cells (SE) in Balb/c mice, when given as from 1–2 days prior to the antigen injection. A slight stimulation of the immune response was observed when levan was given shortly before or 1 day after the antigen.