检索结果-南通市图书馆

Discriminative graph regularized broad learning system for image recognition

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2018年第11期61卷 179-192页

作者： Junwei JIN zhulin liu C.L.Philip CHEN Faculty of Science and Technology University of Macau Dalian Maritime University State Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of Sciences

Broad learning system(BLS) has been proposed as an alternative method of deep learning. The architecture of BLS is that the input is randomly mapped into series of feature spaces which form the feature nodes, and the output of the feature nodes are expanded broadly to form the enhancement nodes, and then the output weights of the network can be determined analytically. The most advantage of BLS is that it can be learned incrementally without a retraining process when there comes new input data or neural nodes. It has been proven that BLS can overcome the inadequacies caused by training a large number of parameters in gradient-based deep learning algorithms. In this paper, a novel variant graph regularized broad learning system(GBLS) is proposed. Taking account of the locally invariant property of data, which means the similar images may share similar properties, the manifold learning is incorporated into the objective function of the standard BLS. In GBLS, the output weights are constrained to learn more discriminative information,and the classification ability can be further enhanced. Several experiments are carried out to verify that our proposed GBLS model can outperform the standard BLS. What is more, the GBLS also performs better compared with other state-of-the-art image recognition methods in several image databases.

关键词： broad learning system deep learning graph regularization image recognition feature extraction incremental learning

Broad learning system based on the quantized minimum error entropy criterion

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2022年第12期65卷 204-219页

作者： Simin ZHANG zhulin liu C.L.Philip CHEN School of Computer Science and Engineering South China University of Technology Pazhou Lab

The broad learning system(BLS) based on the minimum mean square error(MMSE) criterion can achieve outstanding performance without spending too much time in various machine learning ***, when data are polluted by non-Gaussian noise, the stability of BLS may be destroyed because the MMSE criterion is sensitive to outliers. Different from the MMSE criterion, the minimum error entropy(MEE) criterion utilizes the kernel function to capture high-dimensional information and decrease the negative influence of outliers, which will make BLS more discriminative and robust. Nevertheless, the computational complexity of MEE is high due to a double summation of the data size. To solve these issues, this paper proposes a new robust BLS variant based on the quantized minimum error entropy(QMEE) criterion, in which a quantization operation is used to reduce the computational complexity of MEE. The proposed model BLS-QMEE is optimized by the fixed-point iterative method, and a sufficient condition for its convergence is provided. Compared with the standard BLS and other existing robust variants of BLS, BLS-QMEE performs more satisfactorily without consuming too much time. The desirable performance of BLS-QMEE is verified by various experiments on function approximation, several public datasets, and a practical application.

关键词： broad learning system quantized minimum error entropy robustness minimum error entropy convergence

Low band-gap benzodithiophene-thienothiophenecopolymers: the effect of dual two-dimensional substitutions on optoelectronic properties

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China Chemistry 2015年第2期58卷 267-275页

作者： zhulin liu Jiangman Sun Yongxiang Zhu Peng liu Lianjie Zhang Junwu Chen Fei Huang Yong Cao Institute of Polymer Optoelectronic Materials and Devices State Key Laboratory of Luminescent Materials and DevicesSouth China University of Technology

Two new conjugated copolymers, PBDT-T6-TTF and PBDT-T12-TTF, were derived from a novel 4-fluorobenzoyl thienothi- ophene （TTF）. In addition, two types of benzodithiophene （BDT） units with 2,3-dihexylthienyl （T6） and 2,3-didodecylthienyl （T12） substituents, respectively, were successfully synthesized. The effect of the dual two-dimensional （2D） substitutions of the building blocks upon the optoelectronic properties of the polymers was investigated. Generally, the two polymers exhibited good solubility and broad absorption, showing similar optical band gaps of ~1.53 eV. However, PBDT-T6-TTF with its shorter alkyl chain length possessed a larger extinction coefficient in thin solid film. The highest occupied molecular orbital （HOMO） level of PBDT-T6-TTF was located at -5.38 eV while that of PBDT-T12-TTF was at -5.51 eV. In space charge-limited- current （SCLC） measurement, PBDT-T6-TTF and PBDT-T12-TTF displayed respective hole mobilities of 3.0~10-~ and 1.6x10 5 cm2 V-1 s-l. In polymer solar cells, PBDT-T6-TTF and PBDT-T12-TTF showed respective power conversion efficiencies （PCEs） of 2.86% and 1.67%. When 1,8-diiodooctane （DIO） was used as the solvent additive, the PCE of PBDT-T6-TTF was remarkably elevated to 4.85%, but the use of DIO for the PBDT-T12-TTF-blend film resulted in a lower PCE of 0.91%. Atomic force microscopy （AFM） indicated that the superior efficiency of PBDT-T6-TTF with 3% DIO （v/v） should be related to the better continuous phase separation of the blend film. Nevertheless, the morphology of the PBDT-T12-TTF deteriorated when the 3% DIO （v/v） was added. Our results suggest that the alkyl-chain length on the 2D BDT units play an important role in determining the optoelectronic properties of dual 2D BDT-TT-based polymers.

关键词： conjugated copolymers benzodithiophene thienothiophene dual 2-dimentional substitutions alkyl side-chain length

在线全文

学校读者我要写书评

暂无评论

最优解逼近下配电网线损故障诊断方法设计

粘接 2024年第10期51卷 193-196页

作者：李连浩路鑫杜瑞红王琦刘祝麟国网冀北电力有限公司承德供电公司河北承德067000 国网冀北电力有限公司张家口供电公司河北张家口075000 保定电力职业技术学院河北保定071051 国网冀北电力有限公司隆化县供电公司分公司河北隆化068150 国网冀北电力有限公司丰宁县供电公司河北丰宁068350

针对单纯以一次阈值思路计算引起线损异常导致故障诊断准确率不高问题,设计最优解逼近的配电网线损故障诊断方法。使用等效容量法计算配电变压器容量的分配比值,并根据全网线损功率的变化情况计算接入DG后的配电网线损参数。搭建三层BP... 详细信息

针对单纯以一次阈值思路计算引起线损异常导致故障诊断准确率不高问题,设计最优解逼近的配电网线损故障诊断方法。使用等效容量法计算配电变压器容量的分配比值,并根据全网线损功率的变化情况计算接入DG后的配电网线损参数。搭建三层BP神经网络结构,利用拓扑匹配方法对配电网的异常损耗进行匹配,当判定为异常状态时,通过遗传算法优化BP神经网络的权重和阈值,对线损的异常类型进行标记和定位,得到线损异常点的定位结果,根据定位结果进行线损故障诊断。实验结果表明,使用所提方法对配电网线损故障进行诊断,诊断准确率较高,具有较好的应用价值。

关键词：配电网线损故障诊断遗传算法-多层前馈网络全网线损功率等效容量法

Edaravone-loaded poly(amino acid) nanogel inhibits ferroptosis for neuroprotection in cerebral ischemia injury

在线全文

学校读者我要写书评

暂无评论

Asian Journal of Pharmaceutical Sciences 2024年第2期19卷 89-101页

作者： Yunhan Zhang zhulin Zou Shuang liu Fangfang Chen Minglu Li Haoyang Zou Haiyan liu Jianxun Ding Key Laboratory of Pathobiology Ministry of Education Department of AnatomyCollege of Basic Medical SciencesJilin UniversityChangchun 130061China Key Laboratory of Polymer Ecomaterials Changchun Institute of Applied ChemistryChinese Academy of SciencesChangchun 130022China Department of Gastrointestinal Colorectaland Anal SurgeryChina-Japan Union Hospital of Jilin UniversityChangchun 130033China

Neurological injury caused by ischemic stroke is a major cause of permanent disability and death. The currently available neuroprotective drugs fail to achieve desired therapeutic efficacy mainly due to short circulation half-life and poor blood−brain barrier (BBB) permeability. For that, an edaravone-loaded pH/glutathione (pH/GSH) dual-responsive poly(amino acid) nanogel (NG/EDA) was developed to improve the neuroprotection of EDA. The nanogel was triggered by acidic and EDA-induced high-level GSH microenvironments, which enabled the selective and sustained release of EDA at the site of ischemic injury. NG/EDA exhibited a uniform sub-spherical morphology with a mean hydrodynamic diameter of 112.3 ± 8.2 nm. NG/EDA efficiently accumulated at the cerebral ischemic injury site of permanent middle cerebral artery occlusion (pMCAO) mice, showing an efficient BBB crossing feature. Notably, NG/EDA with 50 µM EDA significantly increased neuron survival (29.3%) following oxygen and glucose deprivation by inhibiting ferroptosis. In addition, administering NG/EDA for 7 d significantly reduced infarct volume to 22.2% ± 7.2% and decreased neurobehavioral scores from 9.0 ± 0.6 to 2.0 ± 0.8. Such a pH/GSH dual-responsive nanoplatform might provide a unique and promising modality for neuroprotection in ischemic stroke and other central nervous system diseases.

关键词： Poly(amino acid)nanogel Controlled drug delivery Inhibition of ferroptosis Neuroprotection Cerebral ischenia injury therapy

Numerical analysis on the transport properties and residence time distribution of ribbon biomass particles in a riser reactor based on CFD-DEM approach

在线全文

学校读者我要写书评

暂无评论

Particuology 2024年第4期87卷 133-146页

作者： Haichao Zhao Conghui Gu Bingyang Xu Yuan liu Mingpu Du Kaiyuan Deng Jingyu Zhu Iana Voronina zhulin Yuan School of Energy and Power Jiangsu University of Science and TechnologyZhenjiang212100China China Tobacco Jiangxi Industrial Co.Ltd. Nanchang330096China Institute of Mechanics and Power Engineering Ogarev Mordovia State UniversitySaransk430005Russia Key Laboratory of Energy Thermal Conversion and Control of Ministry of Education School of Energy and EnvironmentSoutheast UniversityNanjing210096China

A bended ribbon biomass particle model was developed to explore the dynamic transport properties inside a riser *** time distribution(RTD)of the particles was analyzed by using the Eulerian-Lagrange *** effects of sampling height,particle density,particle size and gas-to-solid mass ratio on RTD were *** coupled Computational Fluid Dynamics and Discrete Element Method(CFD-DEM)model was verified firstly by experimental data on pressure drop and residence time distribution density *** simulation results demonstrated that the ribbon biomass particles display a typical annular-core spatial distribution during *** RTD of particles exhibit an approximate single-peaked normal *** mean residence time(MRT)can reach up to 0.7 s when the particle density is 1200 kg/m^(3).Particle with higher density has longer mean residence *** flow patterns are closer to plug flow if particle length over 12 *** particle flow pattern is not sensitive to changes in particle density and size,while the gas-to-material mass ratio has a significant impact on it.

关键词： Riser Gas-solids two-phase flow Ribbon biomass particles Residence time distribution CFD-DEM

CSL学习者词语跨类混淆的分布特征与认知动因

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

河南大学学报（社会科学版） 2022年第1期62卷 112-116,155页

作者：刘竹林北京交通大学语言与传播学院北京100044

词语跨类混淆是CSL学习者的一种普遍性、高发性的混淆现象。词语跨类混淆并不是任意的,在宏观、中观、微观三个层面均存在明显的误用倾向。这表明在CSL学习者的心理词典中,词类范畴具有较强的心理现实性。人类认知模式的临近性及显著度... 详细信息

词语跨类混淆是CSL学习者的一种普遍性、高发性的混淆现象。词语跨类混淆并不是任意的,在宏观、中观、微观三个层面均存在明显的误用倾向。这表明在CSL学习者的心理词典中,词类范畴具有较强的心理现实性。人类认知模式的临近性及显著度特征可以对这些高频、明显的误用倾向作出解释:词语跨类混淆现象本质上是一种错误的转喻。

关键词：词语偏误易混淆词词语跨类混淆汉语习得

在线全文

学校读者我要写书评

暂无评论

氧化锌掺杂改性研究进展

材料导报 2014年第21期28卷 47-52页

作者：夏志美刘竹林湖南工业大学冶金工程学院株洲412007

氧化锌半导体材料既是优良的压电材料,也是性能优异的多功能材料,在太阳能电池、显示器件和气体传感器等电子元器件领域以及光催化降解有机污染物领域的应用受到重视。综述了不同掺杂元素对氧化锌材料的光、电、磁、气敏以及变阻等特性... 详细信息

氧化锌半导体材料既是优良的压电材料,也是性能优异的多功能材料,在太阳能电池、显示器件和气体传感器等电子元器件领域以及光催化降解有机污染物领域的应用受到重视。综述了不同掺杂元素对氧化锌材料的光、电、磁、气敏以及变阻等特性的影响,针对氧化锌掺杂的研究现状,提出了今后的研究方向:研究掺杂原子的选择定则;加强掺杂氧化锌材料的应用研究;深入研究氧化锌掺杂引起半导体特性改变的机理。

关键词：氧化锌掺杂改性光电磁特性气敏性变阻特性

在线全文

学校读者我要写书评

暂无评论

氧化锌的p型掺杂改性研究进展

半导体技术 2014年第8期39卷 561-566页

作者：夏志美刘竹林湖南工业大学冶金工程学院湖南株洲412007

综述了近年来纳米氧化锌材料p型掺杂改性的研究现状,介绍了用于p型掺杂的单元素和共掺杂双元素以及所采用的主要掺杂方法,这些方法包括热蒸发法、喷雾热解法、溶胶-凝胶法、低温水解法、离子束增强沉积法、脉冲激光沉积法和磁控溅射法... 详细信息

综述了近年来纳米氧化锌材料p型掺杂改性的研究现状,介绍了用于p型掺杂的单元素和共掺杂双元素以及所采用的主要掺杂方法,这些方法包括热蒸发法、喷雾热解法、溶胶-凝胶法、低温水解法、离子束增强沉积法、脉冲激光沉积法和磁控溅射法等。探讨了掺杂后的氧化锌在光电磁以及p型稳定性等方面表现出的新特性,介绍了p型氧化锌薄膜在紫外探测器和太阳电池方面的应用,对p型氧化锌薄膜未来的发展进行了分析和预测,认为制备稳定、低电阻率和高载流子浓度的p型氧化锌薄膜是今后研究的重点。

关键词：氧化锌 p型掺杂掺杂元素掺杂方法低电阻率