Nanoparticles from compounds with layered structures

Layered structures of the type B8, C6, and C7 excel in the fact that the binding within the layers between the metal and the non-metal is of the covalent type and between the layers of the van der Waals type. This leads, in the case of nanoparticles, to a large number of dangling bonds on the periphery of the layers. In the well-studied case of graphite, the formation of onion-like structures and nanotubes reduces the free energy of the system. It is shown that the formation of these morphologies is very general for small particles. This is shown for BN, MoS₂, WS₂, WSe₂, and MoSe₂. SnS₂ and ZrSe₂ are special exceptions, as nanoparticles of these compounds preferably form rope-like structures. These experiments lead to the general rule that nanoparticles made of layered compounds try to form closed structures.     

Biophenolic profile in olives by nuclear magnetic resonance

The molecular composition of biophenols in olive fruit was investigated in order to experiment with novel procedures for the determination of these microcomponents in fresh and processed table olives. The presence of biophenols in table olives, with recognized antioxidant activity, can be strictly linked to the texture and the organoleptic characteristics of the food product, giving a functional value to this Mediterranean food. Olives from Spain (“Hojiblanca” cv.), Portugal (“Douro” cv.), Greece (“Conservolia” and “Thasos”cv.) and Italy (“Taggiasca” and “Cassanese” cv.) were examined, because the experimental data, checked by high-performance liquid chromatography (HPLC), show molecular composition differences in the tested samples related to the geographic area of analyzed olive fruit cultivars [Bianco, A., & Uccella, N. (2000). Biophenolic components of olives. Food Research International, 33, 475–485]. Three different protocols were utilized: the first one allows the determination of the biophenolic content present as free and esterified compounds; the second affords the total biophenolic content; the third indicates the biophenols present as glycosides. The biophenolic content, which was previously determined by classic HPLC methods (Bianco & Uccella, 2000), was checked by a simple ¹H-NMR experiment. The comparison between the data obtained from ¹H-NMR with those measured by HPLC, indicates a good agreement and suggests the possibility of employing ¹H-NMR for the rapid determination of biophenolic content in olives and also in other foods.     

阿里云物联网平台的混流泵房远程监控系统设计

针对大型农业园区混流泵房较为分散、可控性和可观测性低等问题,设计并实现了基于阿里云的混流泵房远程监控系统.系统以STM32F429为主控芯片,ATK-M751模块为主要通信模块,利用多种环境监测传感器和霍尔转速传感器实时采集混流泵房的环境以及混流泵转速.系统通过MQTT协议实现设备与阿里云物联网平台的数据交互.用户能够...     

轻度认知障碍的BOLD信号MEMD降噪处理

早期诊断轻度认知障碍是干预阿尔茨海默症的有效途径.目前常使用静息态功能磁共振成像和机器学习方法进行轻度认知障碍的辅助诊断,其关键是使用血氧水平依赖(blood oxygenation level dependent, BOLD)信号构建大脑的功能性连接.针对大脑静息态BOLD信号中存在各种外界噪音干扰的问题,提出结合多元经验模态分解与皮尔逊相关的重构方法与极正极负重构准则,将大脑默认模式网络的中心节点后扣带回皮层作为模板,重构BOLD信号以降低外界噪音干扰.实验结果表明,基于极正极负重构准则降噪后的BOLD信号构建功能性连接,相较降噪前的数据,在分类性能方面可以提高数据的差异性,在特征选择性能方面可以对数据集降维的同时进一步提升分类性能.此外,以上性能均优于传统重构准则.最后,对降噪后的最优特征子集进行统计性分析,发现脑岛可能是默认模式网络的相关脑区,小脑蚓体与后扣带回皮层可能构成一种认知功能补偿网络,这是以往研究中少有提出的结论.     

工业噪声环境下多麦状态空间模型语音增强算法

在协同作业的工业环境中进行语音通信时,语音往往会淹没于工业噪声中,致使语音通信的有效性受到影响。针对这种工业噪声下的语音环境,提出了一种采用多麦克风的卡尔曼语音增强算法。该算法简化了状态空间模型(SSM)中的差分方程以降低复杂度,每个采样点实时得到去噪信号从而增强了实时性。另外,为了进一步简化复杂度,还利用最小二乘原则来对语音进行增强。实验中采用了公开数据库的语音信号和工厂噪声信号来模拟多麦下的带噪语音,将所提算法与传统算法进行了对比。实验结果表明,所提算法的输出语噪比(增强后的语音与残留噪声之比)优于传统算法约2 dB,而运行时间仅不到传统算法的2%,且延迟时间仅是毫秒级。     

Phenolic composition,antioxidant and cytoprotective effects of aqueous-methanol extract from Anneslea fragrans leaves as affected by drying methods

The present study is aimed to evaluate the effect of three different drying methods, including shade drying (SD), freeze drying (FD) and hot air dying (HAD), on the chemical components, antioxidant and cytoprotective activities of aqueous-methanol extracts from A. fragrans leaves. The total phenolic and flavonoid content were determined by spectroscopic methods, whereas the chemical profile of different extracts was determined by ultra-high-performance liquid chromatography coupled with a tandem mass spectrometry (UHPLC-MS/MS). A total of thirteen compounds were characterised and 9 phenolic compounds have been quantified. Two dihydrochalcone glycosides named confusoside and vacciniifolin were found to be the most abundant phenolic compounds in the SD extract of A. fragrans leaves, which also showed the strongest antioxidant and cytoprotective activities. The results showed that shade drying is an efficient method for preserving the chemical constituents and maintaining the antioxidant and cytoprotective activities of A. fragrans leaves.     

Co掺杂尖晶石型LiMn_2O_4正极材料研究进展

尖晶石型LiMn_2O_4具有资源广、电压高、成本低、无污染、安全等优点,是具有良好发展前景的锂离子电池正极材料之一。由于在使用过程中Jahn-Teller效应、锰的溶解等引起其结构塌陷,制约了材料的应用,掺杂可以有效稳定LiMn_2O_4材料的结构、抑制容量衰减等,提高循环稳定性能。综述了Co掺杂尖晶石型LiMn_2O_4正极材料的研究进展,展望了Co掺杂尖晶石型LiMn_2O_4材料的发展前景。     

RFID标签冲突分离的最大后验概率聚类

无线射频识别（Radio frequency identification,RFID）通信系统中,当多个标签冲突时,可对冲突信号分离再解码以提高系统的通信效率,而信号分离通常依靠聚类完成。然而,有时传统方法在算法时间复杂度和聚类准确度方面不可兼顾。本文提出一种最大后验概率估计的聚类方法,通过蒙特卡罗方法快速找到峰值,以此作为聚类中心完成信号分离。实验分别采用了仿真和软件无线电的实测数据对算法进行测试,两者的结果均表明,本文算法在高信噪比下具有较高的聚类准确度和较低的时间复杂度,且嵌入动态帧ALOHA系统中,吞吐量可达到0.55,高于纯动态帧ALOHA。     

轻度认知障碍磁共振信号中固有频率动态功能性连接的聚类研究

功能性连接（Functional connectivity, FC）可以表示脑区的协同工作能力,目前广泛采用动态功能性连接（Dynamic functional connectivity,DFC）和聚类分析相结合的方法研究疾病的显著性差异分析和分类。但现有研究,对于聚类个数的确定和聚类结果选用并没有明确的标准,且传统的DFC无法表示不同频率的FC信息。因此,本文对轻度认知障碍（Mild cognitive impairment, MCI）磁共振信号中固有频率DFC聚类问题进行研究。首先对被试的时间进程（Time course,TC）数据做噪音辅助的多元经验模态分解并计算DFC;然后通过评判辅助的聚类方法做聚类分析,再采用最小二乘对聚类结果做拟合;最后采用分类器做分类。实验采用阿尔茨海默病神经影像学（Alzheimer’s disease nearoimaging, ADNI）数据库的数据对本文算法进行测试。实验结果表明,有监督聚类分类准确率高于无监督聚类;引入固有频率的DFC分类准确率要高于传统的DFC;最小二乘拟合能提升分类准确率。