资源类型

期刊论文 80

会议视频 1

年份

2023 10

2022 4

2021 8

2020 7

2019 10

2018 7

2017 3

2016 3

2015 5

2014 1

2013 5

2012 1

2011 2

2009 3

2008 2

2007 1

2006 3

2004 1

2002 2

2001 1

展开 ︾

关键词

气候变化 2

知识工程 2

频域分析 2

2D—3D配准 1

Maradbcm算法 1

PID控制 1

P波 1

QRS-T复合波 1

东亚季风 1

中断型协调器 1

主题模型 1

主题爬虫；本体；优先度评估；模拟退火；暴雨灾害 1

习惯领域 1

互相关 1

人- 信息- 物理系统（HCPS） 1

人- 物理系统（HPS） 1

人为地形变化 1

人群计数；知识蒸馏；终身学习 1

代谢 1

展开 ︾

检索范围：

排序： 展示方式：

无监督域自适应的动态参数化学习 Research Article

蒋润华1,2,韩亚洪1,2

《信息与电子工程前沿（英文）》 2023年 第24卷 第11期   页码 1616-1632 doi: 10.1631/FITEE.2200631

摘要： 无监督领域自适应通过学习域不变表示实现神经网络从有标签数据组成的源域到无标签数据组成的目标域迁移。近期研究通过直接匹配这两个域的边缘分布实现这一目标。然而，已有研究大多数忽略域对齐和语义判别学习之间的动态平衡，因此容易受负迁移和异常样本影响。为解决这些问题，引入动态参数化学习框架。首先，通过探索领域级语义知识，提出动态对齐参数自适应地调整域对齐和语义判别学习的优化过程。此外，为获得判别能力强和域不变的表示，提出在源域和目标域上对齐优化过程。本文通过综合实验证明了所提出方法的有效性，并在3个视觉任务的7个数据集上进行广泛比较，证明可行性。

关键词： 无监督领域自适应；优化步骤；跨域判别表示；语义判别    

HTML PDF 收藏

双向协同的去中心化多源域自适应 Research Article

魏义康1,2,韩亚洪1,2

《信息与电子工程前沿（英文）》 2022年 第23卷 第12期   页码 1780-1794 doi: 10.1631/FITEE.2200284

摘要： 去中心化多源域自适应是指在数据去中心化场景下执行无监督多源域自适应。数据去中心化的挑战是源域与目标域在训练中缺乏跨域协同。对于无标签的目标域，目标域模型需要在源域模型的协助下迁移监督知识，而域差距会导致源域模型的适应性能有限。对于有标签的源域，源域模型在数据去中心化场景下倾向于过拟合本地数据，从而导致负迁移问题。对于以上挑战，提出双向协同的去中心化多源域自适应方法，通过其它域模型的协助进行局部源域模型与局部目标域模型的协同训练与聚合。对于目标域，我们在源域模型的协助下蒸馏监督知识，同时完全利用无标签目标域的数据来缓解域偏移问题。对于源域，我们在目标域模型的协助下正则化源域模型来避免负迁移问题。以上过程在去中心化的源域和目标域之间形成一种双向协同，以便在数据去中心化场景下提升域自适应性能。在标准多源域自适应数据集上的实验表明，我们的方法以较大优势优于现有的多源域自适应方法。

关键词： 多源域自适应；数据去中心化；域偏移；负迁移    

HTML PDF 收藏

针对无监督域自适应问题的深度逐层领域修正算法 Article

Shuang LI, Shi-ji SONG, Cheng WU

《信息与电子工程前沿（英文）》 2018年 第19卷 第1期   页码 91-103 doi: 10.1631/FITEE.1700774

摘要： 为借助深度网络解决领域偏移问题，本文提出逐层领域修正（layer-wise domain correction, LDC）深度域自适应算法。

关键词： 无监督域自适应；最大均值偏差；残差网络；深度学习    

HTML PDF 收藏

Automatically building large-scale named entity recognition corpora from Chinese Wikipedia

Jie ZHOU,Bi-cheng LI,Gang CHEN

《信息与电子工程前沿（英文）》 2015年 第16卷 第11期   页码 940-956 doi: 10.1631/FITEE.1500067

摘要： Named entity recognition (NER) is a core component in many natural language processing applications. Most NER systems rely on supervised machine learning methods, which depend on time-consuming and expensive annotations in different languages and domains. This paper presents a method for automatically building silver-standard NER corpora from Chinese Wikipedia. We refine novel and language-dependent features by exploiting the text and structure of Chinese Wikipedia. To reduce tagging errors caused by entity classification, we design four types of heuristic rules based on the characteristics of Chinese Wikipedia and train a supervised NE classifier, and a combined method is used to improve the precision and coverage. Then, we realize type identification of implicit mention by using boundary information of outgoing links. By selecting the sentences related with the domains of test data, we can train better NER models. In the experiments, large-scale NER corpora containing 2.3 million sentences are built from Chinese Wikipedia. The results show the effectiveness of automatically annotated corpora, and the trained NER models achieve the best performance when combining our silver-standard corpora with gold-standard corpora.

关键词： NER corpora     Chinese Wikipedia     Entity classification     Domain adaptation     Corpus selection    

HTML PDF 收藏

Nonlinear design-point performance adaptation approaches and their comparisons for gas turbine applications

Y. G. LI, P. PILIDIS,

《能源前沿（英文）》 2009年 第3卷 第4期   页码 446-455 doi: 10.1007/s11708-009-0042-9

摘要： Accurate performance simulation and understanding of gas turbine engines is very useful for gas turbine manufacturers and users alike and such a simulation normally starts from its design point. When some of the engine component parameters for an existing engine are not available, they must be estimated in order that the performance analysis can be started. Therefore, the simulated design point performance of an engine may be slightly different from its actual performance. In this paper, two nonlinear gas turbine design-point performance adaptation approaches have been presented to best estimate the unknown component parameters and match available design point engine performance, one using a nonlinear matrix inverse adaptation method and the other using a Genetic Algorithm-based adaptation approach. The advantages and disadvantages of the two adaptation methods have been compared with each other. In the approaches, the component parameters may be compressor pressure ratios and efficiencies, turbine entry temperature, turbine efficiencies, engine mass flow rate, cooling flows, and by-pass ratio, etc. The engine performance parameters may be thrust and SFC for aero engines, shaft power, and thermal efficiency for industrial engines, gas path pressures, temperatures, etc. To select the most appropriate to-be-adapted component parameters, a sensitivity bar chart is used to analyze the sensitivity of all potential component parameters against the engine performance parameters. The two adaptation approaches have been applied to a model gas turbine engine. The application shows that the sensitivity bar chart is very useful in the selection of the to-be-adapted component parameters, and both adaptation approaches are able to produce good quality engine models at design point. The comparison of the two adaptation methods shows that the nonlinear matrix inverse method is faster and more accurate, while the genetic algorithm-based adaptation method is more robust but slower. Theoretically, both adaptation methods can be extended to other gas turbine engine performance modelling applications.

关键词： gas turbine     engine     performance adaptation     performance matching     design-point performance simulation     influence coefficient matrix     genetic algorithm    

HTML PDF 收藏

Time-domain and frequency-domain approaches to identification of bridge flutter derivatives

Zhengqing CHEN

《结构与土木工程前沿（英文）》 2009年 第3卷 第2期   页码 173-179 doi: 10.1007/s11709-009-0034-1

摘要： Flutter derivatives are essential for flutter analysis of long-span bridges, and they are generally identified from the vibration testing data of a sectional model suspended in a wind tunnel. Making use of the forced vibration testing data of three sectional models, namely, a thin-plate model, a nearly streamlined model, and a bluff-body model, a comparative study was made to identify the flutter derivatives of each model by using a time-domain method and a frequency-domain method. It was shown that all the flutter derivatives of the thin-plate model identified with the frequency-domain method and time-domain method, respectively, agree very well. Moreover, some of the flutter derivatives of each of the other two models identified with the two methods deviate to some extent. More precisely, the frequency-domain method usually results in smooth curves of the flutter derivatives. The formulation of time-domain method makes the identification results of flutter derivatives relatively sensitive to the signal phase lag between vibration state vector and aerodynamic forces and also prone to be disturbed by noise and nonlinearity.

关键词： long-span bridges     wind-induced vibration     flutter derivatives     forced vibration test     time-domain method     frequency-domain method    

HTML PDF 收藏

Entity and relation extraction with rule-guided dictionary as domain knowledge

《工程管理前沿（英文）》   页码 610-622 doi: 10.1007/s42524-022-0226-0

摘要： Entity and relation extraction is an indispensable part of domain knowledge graph construction, which can serve relevant knowledge needs in a specific domain, such as providing support for product research, sales, risk control, and domain hotspot analysis. The existing entity and relation extraction methods that depend on pretrained models have shown promising performance on open datasets. However, the performance of these methods degrades when they face domain-specific datasets. Entity extraction models treat characters as basic semantic units while ignoring known character dependency in specific domains. Relation extraction is based on the hypothesis that the relations hidden in sentences are unified, thereby neglecting that relations may be diverse in different entity tuples. To address the problems above, this paper first introduced prior knowledge composed of domain dictionaries to enhance characters’ dependence. Second, domain rules were built to eliminate noise in entity relations and promote potential entity relation extraction. Finally, experiments were designed to verify the effectiveness of our proposed methods. Experimental results on two domains, including laser industry and unmanned ship, showed the superiority of our methods. The F1 value on laser industry entity, unmanned ship entity, laser industry relation, and unmanned ship relation datasets is improved by +1%, +6%, +2%, and +1%, respectively. In addition, the extraction accuracy of entity relation triplet reaches 83% and 76% on laser industry entity pair and unmanned ship entity pair datasets, respectively.

关键词： entity extraction     relation extraction     prior knowledge     domain rule    

HTML PDF 收藏

Damage identification in connections of moment frames using time domain responses and an optimization

《结构与土木工程前沿（英文）》 2021年 第15卷 第4期   页码 851-866 doi: 10.1007/s11709-021-0739-3

摘要： Damage is defined as changes to the material and/or geometric properties of a structural system, comprising changes to the boundary conditions and system connectivity, adversely affecting the system’s performance. Inspecting the elements of structures, particularly critical components, is vital to evaluate the structural lifespan and safety. In this study, an optimization-based method for joint damage identification of moment frames using the time-domain responses is introduced. The beam-to-column connection in a metallic moment frame structure is modeled by a zero-length rotational spring at both ends of the beam element. For each connection, an end-fixity factor is specified, which changes between 0 and 1. Then, the problem of joint damage identification is converted to a standard optimization problem. An objective function is defined using the nodal point accelerations extracted from the damaged structure and an analytical model of the structure in which the nodal accelerations are obtained using the Newmark procedure. The optimization problem is solved by an improved differential evolution algorithm (IDEA) for identifying the location and severity of the damage. To assess the capability of the proposed method, two numerical examples via different damage scenarios are considered. Then, a comparison between the proposed method and the existing damage identification method is provided. The outcomes reveal the high efficiency of the proposed method for finding the severity and location of joint damage considering noise effects.

关键词： damage identification     beam-to-column connection     time-domain response     optimization    

HTML PDF 收藏

Terahertz time-domain spectroscopy of high-pressure flames

Jason BASSI, Mark STRINGER, Bob MILES, Yang ZHANG

《能源前沿（英文）》 2009年 第3卷 第2期   页码 123-133 doi: 10.1007/s11708-009-0033-x

摘要： Laser spectroscopy in the visible and near infrared is widely used as a diagnostic tool for combustion devices, but this approach is difficult at high pressures within a sooty flame itself. High soot concentrations render flames opaque to visible light, but they remain transparent to far-infrared or terahertz (THz) radiation. The first far-infrared absorption spectra, to the best of our knowledge, of sooty, non-premixed, ethylene high-pressure flames covering the region of 0.2-2.5 THz is presented. A specially designed high-pressure burner which is optically accessible to THz radiation has been built allowing flame transmission measurements up to pressures of 1.6 MPa. Calculations of the theoretical combustion species absorption spectra in the 0.2-3 THz range have shown that almost all the observable features arise from H O. A few OH (1.84 and 2.51 THz), CH (2.58 THz), and NH (1.77 and 2.95 THz) absorption lines are also observable in principle. A large number of H O absorption lines are observed in the ground vibrational in a laminar non-premixed, sooty flame (ethylene) at pressures up to 1.6 MPa.

关键词： terahertz time-domain spectroscopy     high-pressure flames     H2O absorption lines    

HTML PDF 收藏

framework for underground structures in layered ground under inclined P-SV waves using stiffness matrix and domain

《结构与土木工程前沿（英文）》 2023年 第17卷 第1期   页码 10-24 doi: 10.1007/s11709-022-0904-3

摘要： A numerical framework was proposed for the seismic analysis of underground structures in layered ground under inclined P-SV waves. The free-field responses are first obtained using the stiffness matrix method based on plane-wave assumptions. Then, the domain reduction method was employed to reproduce the wavefield in the numerical model of the soil–structure system. The proposed numerical framework was verified by providing comparisons with analytical solutions for cases involving free-field responses of homogeneous ground, layered ground, and pressure-dependent heterogeneous ground, as well as for an example of a soil–structure interaction simulation. Compared with the viscous and viscous-spring boundary methods adopted in previous studies, the proposed framework exhibits the advantage of incorporating oblique incident waves in a nonlinear heterogeneous ground. Numerical results show that SV-waves are more destructive to underground structures than P-waves, and the responses of underground structures are significantly affected by the incident angles.

关键词： underground structures     seismic response     stiffness matrix method     domain reduction method     P-SV waves    

HTML PDF 收藏

Comments on “Adaptation of Chinese and German maize-based food-feed-energy systems to limited phosphate

null

《农业科学与工程前沿（英文）》 2019年 第6卷 第4期   页码 443-444 doi: 10.15302/J-FASE-2019288

HTML PDF 收藏

Frequency domain a9ctive vibration control of a flexible plate based on neural networks

Jinxin LIU, Xuefeng CHEN, Zhengjia HE

《机械工程前沿（英文）》 2013年 第8卷 第2期   页码 109-117 doi: 10.1007/s11465-013-0252-z

摘要：

A neural-network (NN)-based active control system was proposed to reduce the low frequency noise radiation of the simply supported flexible plate. Feedback control system was built, in which neural network controller (NNC) and neural network identifier (NNI) were applied. Multi-frequency control in frequency domain was achieved by simulation through the NN-based control systems. A pre-testing experiment of the control system on a real simply supported plate was conducted. The NN-based control algorithm was shown to perform effectively. These works lay a solid foundation for the active vibration control of mechanical structures.

关键词： active vibration control (AVC)     neural network (NN)     low frequency noise     frequency domain control     multi-frequency control    

HTML PDF 收藏

Thermal reffusivity: uncovering phonon behavior, structural defects, and domain size

Yangsu XIE, Bowen ZHU, Jing LIU, Zaoli XU, Xinwei WANG

《能源前沿（英文）》 2018年 第12卷 第1期   页码 143-157 doi: 10.1007/s11708-018-0520-z

摘要： To understand the relation between different nanostructures and thermal properties, a simple yet effective model is in demand for characterizing the underlying phonons and electrons scattering mechanisms. Herein, we make a systematic review on the newly developed thermal reffusivity theory. Like electrical resistivity which has been historically used as a theory for analyzing structural domain size and defect levels of metals, the thermal reffusivity can also uncover phonon behavior, structure defects and domain size of materials. We highlight that this new theory can be used for not only metals, but also nonmetals, even for amorphous materials. From the thermal reffusivity against temperature curves, the Debye temperature of the material and the ideal thermal diffusivity of single perfect crystal can be evaluated. From the residual thermal reffusivity at the 0 K limit, the structural thermal domain (STD) size of crystalline and amorphous materials can be obtained. The difference of white hair and normal black hair from heat conduction perspective is reported for the first time. Loss of melanin results in a worse thermal protection and a larger STD size in the white hair. By reviewing the different variation of thermal reffusivity against decreasing temperature profiles, we conclude that they reflected the structural connection in the materials. Ultimately, the future application of thermal reffusivity theory in studying 2D materials and amorphous materials is discussed.

关键词： thermal reffusivity theory     phonon behavior     structure defects     structural thermal domain (STD) size     2D material     amorphous material    

HTML PDF 收藏

FERM domain-containing protein FRMD6 activates the mTOR signaling pathway and promotes lung cancer progression

《医学前沿（英文）》 2023年 第17卷 第4期   页码 714-728 doi: 10.1007/s11684-022-0959-5

摘要： FRMD6, a member of the 4.1 ezrin–radixin–moesin domain-containing protein family, has been reported to inhibit tumor progression in multiple cancers. Here, we demonstrate the involvement of FRMD6 in lung cancer progression. We find that FRMD6 is overexpressed in lung cancer tissues relative to in normal lung tissues. In addition, the enhanced expression of FRMD6 is associated with poor outcomes in patients with lung squamous cell carcinoma (n = 75, P = 0.0054) and lung adenocarcinoma (n = 94, P = 0.0330). Cell migration and proliferation in vitro and tumor formation in vivo are promoted by FRMD6 but are suppressed by the depletion of FRMD6. Mechanistically, FRMD6 interacts and colocalizes with mTOR and S6K, which are the key molecules of the mTOR signaling pathway. FRMD6 markedly enhances the interaction between mTOR and S6K, subsequently increasing the levels of endogenous pS6K and downstream pS6 in lung cancer cells. Furthermore, knocking out FRMD6 inhibits the activation of the mTOR signaling pathway in Frmd6−/− gene KO MEFs and mice. Altogether, our results show that FRMD6 contributes to lung cancer progression by activating the mTOR signaling pathway.

关键词： FRMD6     lung cancer     mTOR pathway    

HTML PDF 收藏

Analytical method of capsizing probability in the time domain for ships in the random beam seas

LIU Liqin, TANG Yougang, LI Hongxia

《结构与土木工程前沿（英文）》 2007年 第1卷 第3期   页码 361-366 doi: 10.1007/s11709-007-0048-5

摘要： The methods for constructing safe basins of ships and predicting their survival probability in random waves were studied. The nonlinear differential equation of the rolling motion of ships in random beam seas was established considering nonlinear damping, nonlinear restoring moment, and random waves. The random rolling differential equation was solved in the time domain by applying the harmonic acceleration method and by synthetically considering the instantaneous state of ships and the narrowband wave energy spectrum. The numerical simulation of random capsizing course was brought forward, the safe basins were constructed for safe navigation, and the survival probabilities of ships were calculated. As an example, the safe basins on the rolling initial value plane were constructed for a 30.27-meter-long fishing vessel according to different initial conditions and random wave parameters. The survival probabilities of the fishing vessel under different significant wave heights were predicted. Thus, the survival probabilities of ships in random seas can be predicted quantitatively by the proposed method.

关键词： different     survival probability     different significant     nonlinear differential     narrowband    

HTML PDF 收藏

标题 作者 时间 类型 操作

无监督域自适应的动态参数化学习

蒋润华1,2,韩亚洪1,2

期刊论文

双向协同的去中心化多源域自适应

魏义康1,2,韩亚洪1,2

期刊论文

针对无监督域自适应问题的深度逐层领域修正算法

Shuang LI, Shi-ji SONG, Cheng WU

期刊论文

Automatically building large-scale named entity recognition corpora from Chinese Wikipedia

Jie ZHOU,Bi-cheng LI,Gang CHEN

期刊论文

Nonlinear design-point performance adaptation approaches and their comparisons for gas turbine applications

Y. G. LI, P. PILIDIS,

期刊论文

Time-domain and frequency-domain approaches to identification of bridge flutter derivatives

Zhengqing CHEN

期刊论文

Entity and relation extraction with rule-guided dictionary as domain knowledge

期刊论文

Damage identification in connections of moment frames using time domain responses and an optimization

期刊论文

Terahertz time-domain spectroscopy of high-pressure flames

Jason BASSI, Mark STRINGER, Bob MILES, Yang ZHANG

期刊论文

framework for underground structures in layered ground under inclined P-SV waves using stiffness matrix and domain

期刊论文

Comments on “Adaptation of Chinese and German maize-based food-feed-energy systems to limited phosphate

null

期刊论文

Frequency domain a9ctive vibration control of a flexible plate based on neural networks

Jinxin LIU, Xuefeng CHEN, Zhengjia HE

期刊论文

Thermal reffusivity: uncovering phonon behavior, structural defects, and domain size

Yangsu XIE, Bowen ZHU, Jing LIU, Zaoli XU, Xinwei WANG

期刊论文

FERM domain-containing protein FRMD6 activates the mTOR signaling pathway and promotes lung cancer progression

期刊论文

Analytical method of capsizing probability in the time domain for ships in the random beam seas

LIU Liqin, TANG Yougang, LI Hongxia

期刊论文