基于变分模态分解和改进粒子群算法优化最小二乘支持向量机的短期电价预测

针对电价序列具有非线性强、波动性大的特点,提出一种基于变分模态分解(VMD)和改进粒子群算法(PSO)优化最小二乘支持向量机(LSSVM)的短期电价预测模型.首先利用VMD将原始电价数据分解为多个子序列,然后采取LSSVM模型分别对子序列进行预测.对于LSSVM预测模型的最优参数选择问题,利用改进的PSO优化LSSVM模型的参数,能够很好地提高模型预测精度.最后集成各子序列预测结果,获得最终电价预测值.为了验证所提模型的有效性,以美国PJM市场电价数据为例进行分析,并与其他预测模型进行对比,结果表明,所提模型能够很好地对短期电价进行预测.     

Synergistic effects of organically modified montmorillonite in combination with metal oxides on the fire safety enhancement of intumescent flame-retarded epoxy resins

Synergistic effects of organically modified montmorillonite (OMMT) in combination with different metal oxides (Bi₂O₃, Sb₂O₃, and MoO₃) on the fire safety enhancement of the intumescent flame-retarded epoxy resins (EPs) were systematically evaluated. The results from X-ray diffraction (XRD) and scanning electron microscopy (SEM) analyses show that the OMMT and metal oxides acquire a uniform distribution in the EP matrix, and OMMT platelets exhibit a fully exfoliated state. The flammability and thermogravimetry (TG) tests show that the intumescent flame retardant (IFR)-OMMT-metal oxide ternary system can endow EPs with the higher synergistic efficiencies on the enhancement of flame retardancy, smoke suppression properties, and charring ability compared to those of IFR or IFR-OMMT system, and the synergistic efficiency is the following order: IFR/OMMT/Sb₂O₃ > IFR/OMMT/MoO₃ > IFR/OMMT/Bi₂O₃. In particular, the sample, filled with 1.5 wt% OMMT, 1.5 wt% Sb₂O₃, and 27 wt% IFR, passes the UL94 V-0 rating and acquires the highest limiting oxygen index value of 28.5% among the samples. The IFR-OMMT-metal oxide ternary system exerts a better synergistic effect on the generation of crosslinking and aromatic structures that supply the excellent charring effect and barrier effect for the EPs, and the synergistic efficiency of IFR-OMMT-metal oxide ternary system is varied with the types of metal oxides.     

Enhanced Photoelectrochemical Performance of Hematite Photoanode by Decorating NiCoP Nanoparticles Through a Facile Spin Coating Method

Catalysis Letters - Hematite (α-Fe2O3) is a potential photoanode material for photoelectrochemical (PEC) water splitting, but its short hole diffusion length and low water oxidation kinetics...     

OsMre11 Is Required for Mitosis during Rice Growth and Development

Meiotic recombination 11 (Mre11) is a relatively conserved nuclease in various species. Mre11 plays important roles in meiosis and DNA damage repair in yeast, humans and Arabidopsis, but little research has been done on mitotic DNA replication and repair in rice. Here, it was found that Mre11 was an extensively expressed gene among the various tissues and organs of rice, and loss-of-function of Mre11 resulted in severe defects of vegetative and reproductive growth, including dwarf plants, abnormally developed male and female gametes, and completely abortive seeds. The decreased number of cells in the apical meristem and the appearance of chromosomal fragments and bridges during the mitotic cell cycle in rice mre11 mutant roots revealed an essential role of OsMre11. Further research showed that DNA replication was suppressed, and a large number of DNA strand breaks occurred during the mitotic cell cycle of rice mre11 mutants. The expression of OsMre11 was up-regulated with the treatment of hydroxyurea and methyl methanesulfonate. Moreover, OsMre11 could form a complex with OsRad50 and OsNbs1, and they might function together in non-homologous end joining and homologous recombination repair pathways. These results indicated that OsMre11 plays vital roles in DNA replication and damage repair of the mitotic cell cycle, which ensure the development and fertility of rice by maintaining genome stability.     

Synthesis,Micellization, and Surface Activity of Novel Linear-Dendritic Carboxylate Surfactants

Two generations of novel linear-dendritic carboxylate surfactants C₁₈-G₁-(COONa)₂ and C₁₈-G₂-(COONa)₄ have been synthesized by the divergent method and their structures are characterized by ¹H Nuclear Magnetic Resonance and Infrared analysis. The electrical conductivity measurement is used to measure the Krafft temperatures of C₁₈-G₁-(COONa)₂ and C₁₈-G₂-(COONa)₄, which are much smaller than those of the corresponding conventional surfactant sodium stearate. The markedly enhanced solubility of two linear-dendritic surfactants is ascribed to the high hydrophilicity of surfactant headgroups induced by the carboxylate and ester groups. The critical micelle concentration (CMC) values obtained from both the electrical conductivity and surface tension measurements indicate that the micellizations of linear-dendritic surfactants become favorable with the increase in the number of the surfactant headgroup. However, the surface activity parameters including the surface tension at the CMC, maximum surface excess, and minimum surface area reveal that C₁₈-G₁-(COONa)₂ exhibits greater efficiency in absorbing at the air/water interface compared to C₁₈-G₂-(COONa)₄, owing to their different steric repulsions of the surfactant headgroups. In addition, C₁₈-G₁-(COONa)₂ and C₁₈-G₂-(COONa)₄ have higher emulsifying ability than the conventional surfactants sodium stearate and sodium octadecyl sulfate.     

金字塔原型对齐的轻量级小样本语义分割网络

小样本图像语义分割任务是计算机视觉领域一个有挑战性的问题,其目标是利用现有一张或几张带有密集分割注释的图片来预测未见类图像的分割掩码.针对该任务,提出了一个基于金字塔原型对齐的轻量级小样本图像语义分割网络.首先,该网络在MobileNetV2网络的深度可分离卷积和逆残差结构基础上,通过金字塔池化模块进行提取特征,保持高维度和低维度的信息,获得不同尺度的特征.同时通过在支持集原型和查询集之间进行相互对齐,使得网络能够从支持集中学到更多的信息,充分利用支持集的信息进行反馈.基于PASCAL-5i数据集的大量实验结果表明,提出的网络结构的均值在1-way 1-shot和1-way 5-shot上分别为49.5％和56.6％,与先进的主流小样本语义分割网络PANet相比分别对应提高了1.4％和0.9％,网络参数量为3.0 MB,相比PANet减小了11.7 MB,同时浮点计算量显著减少,证明了该网络在小样本图像语义分割中的有效性和高效性.     

A Brief Introduction on Joint Radar and Communication Systems

Joint radar and communication(JRC) technology is gradually becoming an essential approach to alleviating spectral congestion. Radar and communications systems were designed with common spectral and hardware resources to reduce size, improve performance, reduce cost, and decongest the spectrum. Various approaches have been proposed to achieve the coexistence of radar and communication systems. This paper mainly focuses on the research directions of radar communication coexistence(RCC) and dual-function radar communication systems(DFRC) in JRC technology. We summarize and analyze the existing research problems in the JRC era. According to the characteristics and advantages of JRC technology, we highlight several potentials in military and commercial applications.     

Multi-Scale Dilated Convolutional Neural Network for Hyperspectral Image Classification

基于多尺度空洞卷积神经网络的高光谱图像分类

郑姗姗¹,刘文¹,单锐¹,赵静一²,江国乾³,张智⁴

（1. 燕山大学理学院,河北 秦皇岛 066004;2. 燕山大学机械工程学院,河北 秦皇岛 066004;3. 燕山大学电气工程学院,河北 秦皇岛 066004;4. 北京航天研究所,北京 100094）

创新点说明：

1）将图像分割方法——空洞卷积用于卷积神经网络进行高光谱图像分类,提取更加广泛、抽象的图像特征。

2）构建基于多尺度空洞卷积神经网络的高光谱图像分类方法。搭建多尺度聚合结构,在每个通道中使用快捷连接和空洞卷积结构,有效提取图像特征,避免信息丢失。

研究目的：

针对图像信息丢失问题,得到高精度的高光谱图像分类方法。

研究方法：

在Indian Pines和Pavia University数据集上,与4个已有的高光谱图像分类方法进行对比实验,比较OA, AA和Kappa值。

研究结果：

多尺度空洞卷积神经网络在Indian Pines和Pavia University数据集上OA值分别达到了99.58%,99.92%。AA值分别达到了99.57%,99.90%。Kappa分别达到了99.52%,99.89%。

结论：

1）在卷积神经网络中引入空洞卷积,可以有效避免图像信息丢失。

2）多尺度空洞卷积神经网络能提取更佳的判别性特征,实现高分类性能。

关键词：多尺度聚合;空洞卷积;高光谱图像分类;快捷连接