http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Adversarial Complementary Learning for Just Noticeable Difference Estimation
Dong Yu,Jian Jin,Lili Meng,Zhipeng Chen,Huaxiang Zhang 한국인터넷정보학회 2024 KSII Transactions on Internet and Information Syst Vol.18 No.2
Recently, many unsupervised learning-based models have emerged for Just Noticeable Difference (JND) estimation, demonstrating remarkable improvements in accuracy. However, these models suffer from a significant drawback is that their heavy reliance on handcrafted priors for guidance. This restricts the information for estimating JND simply extracted from regions that are highly related to handcrafted priors, while information from the rest of the regions is disregarded, thus limiting the accuracy of JND estimation. To address such issue, on the one hand, we extract the information for estimating JND in an Adversarial Complementary Learning (ACoL) way and propose an ACoL-JND network to estimate the JND by comprehensively considering the handcrafted priors-related regions and non-related regions. On the other hand, to make the handcrafted priors richer, we take two additional priors that are highly related to JND modeling into account, i.e., Patterned Masking (PM) and Contrast Masking (CM). Experimental results demonstrate that our proposed model outperforms the existing JND models and achieves state-of-the-art performance in both subjective viewing tests and objective metrics assessments.
Prediction of Damage to Insulation Joints Based on SVM with Unbalanced Data Sets
Dong Yu,Xiao Zi-Qiang 보안공학연구지원센터 2016 International Journal of Multimedia and Ubiquitous Vol.11 No.3
As a key part of track circuit, the state of insulation joints is related to safe, normal and efficient operation of railway. In order to accurately obtain different degrees of insulation joints, a prediction model based on support vector machines has been proposed to study damage to insulation joints. For unbalanced data sets in the research process, a KNN under-sampling is presented to remove redundant and noise samples. By means of BSMOTE over-sampling method to further take full advantage of the data, KNN-BSMOTE-SVM algorithm of hybrid sampling is given to achieve balanced data sets. The theoretical analysis and simulation results show that the proposed algorithm increases classification performance of SVM classifier. Compared with KNN classifier, the classification results of SVM are better, support vector machines used in insulation damaged joints prediction is feasible and effective.
A Fault Diagnosis of ZPW-2000A Tuning Area based on the WNN
Dong Yu,Li Ya-lan,Nan Jie-long 보안공학연구지원센터 2015 International Journal of Multimedia and Ubiquitous Vol.10 No.3
With the rapid development of china railway, ZPW-2000A track circuit has been widely used. Tuning area, which is an important part of the ZPW-2000A track circuit, is not only the relationship between the signal transmission quality, and determine the effect of electrical insulation between adjacent sections. Therefore, the study of fault diagnosis aspects for tuning area is urgent and significant. In this paper, using the theory of transmission line a model of track circuit is built, the comparison of the actual data and experimental data of the track surface voltage envelope curve shows the correctness of this model. Owning to the good time-frequency characteristics of wavelet and the nonlinear mapping features of neural network, a fault diagnosis of ZPW-2000A tuning area based on the wavelet neural network(WNN) is proposed. Combined with the practical failure situation of railway site, the fault diagnosis method in this paper can accurately identify failure modes of tuning area.
A Simulation Study on the Impact of Abandoned Trolley Cases on Pedestrian Evacuation
Dongyue Xu,Jinghong Wang,Juan Liu,Yan Wang,Juncheng Jiang 대한토목학회 2023 KSCE JOURNAL OF CIVIL ENGINEERING Vol.27 No.3
Many pedestrians empirically think that discarding luggage can improve evacuation efficiency, but it is unclear whether discarding luggage is an effective evacuation strategy. Previous models also oversimplified the interference of abandoned luggage on pedestrians. Hence, an extended social force model integrates pedestrian interaction with abandoned luggage like trolley cases was firstly proposed. A simulation based on a real metro station was carried out to examine the effectiveness of evacuation strategy like abandoning trolley cases at the bottleneck such as gate machine and metro carriage door, and the influence mechanism of abandoned trolley cases on pedestrians was investigated. The results show that discarding the trolley cases at the gate machine and at the metro carriage door cannot improve evacuation efficiency but increase the evacuation time by 78.7% and 14.9% at these two positions when the carrying ratio is 50%, respectively. When the occupancy rate of trolley case at the gate machine and metro carriage door is below 19% and 11%, pedestrians tend to detour the abandoned trolley cases and this behavior significantly increased the collisions between pedestrians at bottlenecks, and when the occupancy rate exceeds, they tend to push aside the trolley cases and the evacuation efficiency is further reduced.
Layered Formation-containment Control of Multi-agent Systems in Constrained Space
Dongyu Li,Shuzhi Sam Ge,Guangfu Ma,Wei He 제어·로봇·시스템학회 2020 International Journal of Control, Automation, and Vol.18 No.3
This paper addresses the layered formation-containment (LFC) problem for multiagents in the constrained space with a directed communication topology. The formation-containment problem is first defined using a layered framework, and a layered distributed finite-time estimator is proposed to acquire the target states for agents in each layer. Based on the proposed framework, the formation configuration and the mechanism of the information flow can be explored and designed naturally. To avoid collisions with borders, obstacles, as well as the other agents in the constrained space, an artificial potential function is designed based on the Dirac delta function. Further, a disturbance observer and adaptive neural networks (NNs) are applied to respectively tackle the external disturbance and the model uncertainties. The desired formation of each layer can be achieved while no collision occurs in the constrained space. The semi-global uniform ultimate boundedness of closed-loop errors is guaranteed by Lyapunov stability theory. Simulation results are given to show the effectiveness of the proposed approaches.
Research on Metadata Management in Cloud Computing
Dongyu Li 보안공학연구지원센터 2016 International Journal of Database Theory and Appli Vol.9 No.2
Existing metadata management methods bring about lower access efficiency in solving the problem of renaming directory. This paper proposes a metadata management method based on directory path redirection (named as DPRD) which includes the data distribution method based on directory path and the directory renaming method based on directory path redirection. Experiments show that DPRD effectively solves the lower access efficiency caused by the renaming directory.
The Development and Experimental Test of a New Pore-forming Grouted Precast Shear Wall Connector
Dongyue Wu,Shuting Liang,Zhenxing Guo,Xiaojun Zhu,Qian Fu 대한토목학회 2016 KSCE JOURNAL OF CIVIL ENGINEERING Vol.20 No.4
Pore-forming grouted connection characterized with great economic superiorities and application potentials has weaker mechanical properties comparing to the steel sleeve grouted connector. To improve the mechanical properties of pore-forming grouted connector, the NPGCS was generated using welded closure confinement steels in buckle configuration. Experimental tests and finial element analysis of two NPGCS precast shear wall specimens and one cast-in-situ shear wall specimen was carried out to test mechanical performances. Results indicated that the mechanical properties of NPGCS can match the connection requirements even under the adverse effects from grouted interface gap opening and slip. Contributing to the excessive confinement steels, the mechanical properties of NPGCS specimens are better than that of the cast-in-situ specimen. So, according to the suggestions given out in this paper, the NPGCS can be widely applied into precast buildings in certain areas and conditions.
Electromechanical properties of 2-2 cement based piezoelectric composite
Xu Dongyu,Cheng Xin,Huang Shifeng,Jiang Minhua 한국물리학회 2009 Current Applied Physics Vol.9 No.4
A 2-2 type cement based piezoelectric composites were fabricated by dice-and-fill technique. The influences of lead magnesium niobate-lead zirconate-lead titanate (PMN) ceramic volume fraction on the electromechanical properties of the composite were investigated. The results indicate that the planar electromechanical coupling factor Kp of the composite is hardly influenced by the PMN volume fraction, which fluctuates between 35% and 37%, while both the thickness electromechanical coupling factor Kt and the mechanical quality factor Qm exhibit the trend of increase. When PMN volume fraction is 69.2%, Kt could reach to 49.8%, which is larger than that of the PMN ceramic (Kt = 46%).When PMN volume fraction is 41.7%, Qm is 4.5, which is also much less than that of the PMN ceramic (Qm = 70). This means that the composite has wider frequency band and higher sensitivity to be used as transducers. With the increase of PMN volume fraction, the acoustic impedance of the composite also exhibits the trend of increase. A 2-2 type cement based piezoelectric composites were fabricated by dice-and-fill technique. The influences of lead magnesium niobate-lead zirconate-lead titanate (PMN) ceramic volume fraction on the electromechanical properties of the composite were investigated. The results indicate that the planar electromechanical coupling factor Kp of the composite is hardly influenced by the PMN volume fraction, which fluctuates between 35% and 37%, while both the thickness electromechanical coupling factor Kt and the mechanical quality factor Qm exhibit the trend of increase. When PMN volume fraction is 69.2%, Kt could reach to 49.8%, which is larger than that of the PMN ceramic (Kt = 46%).When PMN volume fraction is 41.7%, Qm is 4.5, which is also much less than that of the PMN ceramic (Qm = 70). This means that the composite has wider frequency band and higher sensitivity to be used as transducers. With the increase of PMN volume fraction, the acoustic impedance of the composite also exhibits the trend of increase.