The structured multiparameter eigenvalue problems in finite element model updating problems

Zhijun Wang,Bo Dong,Yan Yu,Xinzhu Zhao,Yizhou Fang 국제구조공학회 2023 Structural Engineering and Mechanics, An Int'l Jou Vol.88 No.5

The multiparameter eigenvalue method can be used to solve the damped finite element model updating problems. This method transforms the original problems into multiparameter eigenvalue problems. Comparing with the numerical methods based on various optimization methods, a big advantage of this method is that it can provide all possible choices of physical parameters. However, when solving the transformed singular multiparameter eigenvalue problem, the proposed method based on the generalised inverse of a singular matrix has some computational challenges and may fail. In this paper, more details on the transformation from the dynamic model updating problem to the multiparameter eigenvalue problem are presented and the structure of the transformed problem is also exposed. Based on this structure, the rigorous mathematical deduction gives the upper bound of the number of possible choices of the physical parameters, which confirms the singularity of the transformed multiparameter eigenvalue problem. More importantly, we present a row and column compression method to overcome the defect of the proposed numerical method based on the generalised inverse of a singular matrix. Also, two numerical experiments are presented to validate the feasibility and effectiveness of our method.

A Novel Dataset Generating Method for Fine-Grained Vehicle Classification with CNN

Shaoyong Yu,Zhijun Song,Songzhi Su,Wei Li,Yun Wu,Wenhua Zeng 보안공학연구지원센터 2016 International Journal of Database Theory and Appli Vol.9 No.6

We focus on the issue of dataset generation for fine-grained vehicle classification with CNN. Traditionally, to build a large dataset, images must be first collected manually, and then be annotated with a lot of effort. All these work are time-consuming and cost-prohibitive. In this work we propose a novel method that can generate massive images automatically, and these generated images need no annotation. An AutoCAD 3D model of a car of specified make and model is imported into our system, and then images of different views of the car are generated, these images can describe all the details of a car. By taking these images as training dataset, we use a Convolutional Neural Network to train a model for fine-grained vehicle classification. Experimental results show that these images generated virtually by 3D model indeed work as effective as real images.

Energy spread minimization in a cascaded laser wakefield accelerator via velocity bunching

Zhang, Zhijun,Li, Wentao,Liu, Jiansheng,Wang, Wentao,Yu, Changhai,Tian, Ye,Nakajima, Kazuhisa,Deng, Aihua,Qi, Rong,Wang, Cheng,Qin, Zhiyong,Fang, Ming,Liu, Jiaqi,Xia, Changquan,Li, Ruxin,Xu, Zhizhan American Institute of Physics 2016 Physics of plasmas Vol.23 No.5

Detection and Localization of Coordinated State-and-Topology False Data Injection Attack by Multi-modal Learning

Qin Zhijun,Lai Yu 대한전기학회 2022 Journal of Electrical Engineering & Technology Vol.17 No.5

False data injection attack (FDIA) is a typical cyber-attack targeting on power system state estimation. By inserting bias into the power system meter measurements, FDIA can cause errors in state estimation and consequently mislead power system operation. Recently, the coordinated state-and-topology FDIA was developed to falsify both the meter measurements and the topology information to conceal a cyber or physical attack. Conventional machine-learning-based detection methods against FDIA depend on the complete observability of power system topology, therefore cannot detect the coordinated state-andtopology FDIA. To address this challenge, we propose a multi-modal learning model based on Graph Auto-Encoder (GAE) and Residual Neural Networks (ResNet) to detect this type of FDIA. First, GAE is used to learn the compact representation of power system topologies. Then, the obtained topological representation is fused with the measurement to obtain the multi-modal features. Third, ResNet is trained as a multi-label classifi er to accept the fused features and generate an array, which can detect attack scenarios and identify the falsifi ed measurements/topologies by coordinated attacks. Comprehensive case studies using IEEE 9-bus, IEEE 57-bus, and IEEE 118-bus systems are presented. The simulation results show that, as compared to single feature methods, the proposed method is more eff ective in detecting coordinated state-and-topology attacks, and more robust in case of FDIA under unseen topological changes in power system operation

Exploring finger vein based personal authentication for secure IoT

Lu, Yu,Wu, Shiqian,Fang, Zhijun,Xiong, Naixue,Yoon, Sook,Park, Dong Sun North-Holland 2017 Future generations computer systems Vol.77 No.-

<P><B>Abstract</B></P> <P>Personal authentication is getting harder and harder in the internet of things (IoT). Existing methods used for personal authentication, such as passwords and the two-factor authentication (2FA), are inadequate and ineffective due to human error and other attacks. To support more secure IoT, this paper proposes a finger vein based personal authentication method by exploring competitive orientations and magnitudes from finger vein images. Finger vein recognition has been proven to be a reliable and promising solution for biometric-based personal authentication. The stable and rich piecewise line features in finger vein images can be used to clearly represent finger vein patterns for personal authentication. In this paper, we propose an efficient local descriptor for finger vein feature extraction, namely the histogram of competitive orientations and magnitudes (HCOM). For a finger vein image, two types of local histograms are extracted and fused together to efficiently and adequately represent the competitive information: the histogram of competitive orientations (HCO) and the local binary pattern histogram generated from the image of competitive magnitudes (named as HCMLBP). The extensive experimental results from the application of the proposed method to the public finger vein database MMCBNU_6000, demonstrate that the proposed method outperforms state-of-the-art orientation coding (OC)-based methods and other commonly used local descriptors. Additionally, the proposed method can be used for finger vein image enhancement.</P> <P><B>Highlights</B></P> <P> <UL> <LI> The proposed method can efficiently extract competitive orientations and magnitudes. </LI> <LI> The proposed method outperforms the OC-based methods and common local descriptors. </LI> <LI> The proposed method has small feature size and fast speed. </LI> <LI> The proposed method can be used for finger vein image enhancement. </LI> </UL> </P>

Molecular Markers Linked to Specific Characteristics of Prunus persica (L.) Batsch

Mingliang Yu,Ruijuan Ma,Zhijun Shen,Zhen Zhang 한국원예학회 2006 한국원예학회 기타간행물 Vol.- No.-

Paeoniflorin reduces the inflammatory response of THP-1 cells by up‐regulating microRNA-124

Danyun Huang,Zhijun Li,Yue Chen,Yan Fan,Tao Yu 한국유전학회 2021 Genes & Genomics Vol.43 No.6

Background The activation of macrophages and the release of infammatory cytokines are the main reasons for the progress of systemic lupus erythematosus (SLE). MicroRNA (miRNA)-124 is involved in the regulation of macrophages and is a key regulator of infammation and immunity. Objective To explore whether paeoniforin (PF) regulates the biological functions of macrophages depends on miR-124. Methods RT-PCR, WB, ELISA, CCK-8 and fow cytometry were used to evaluate that PF regulated the biological functions of THP-1 cells through miR-124. Results PF signifcantly inhibited the proliferation while promotes the apoptosis of THP-1 cells, and inhibited the release of IL-6, TNF-α and IL-1βin THP-1 cells. RT-PCR results shown that PF up-regulated the expression of miR-124 in THP-1 cells. Functional recovery experiments showed that compared with the LPS+mimic-NC group, LPS+miR-124 mimic signifcantly inhibited the proliferation and the release of IL-6, TNF-α and IL-1β, but promoted the apoptosis of THP-1 cells. In addition, compared with the LPS+PF+inhibitor-NC group, LPS+PF+miR-124 inhibitor signifcantly promoted the proliferation and the release of IL-6, TNF-α and IL-1β, but inhibited the apoptosis of THP-1 cells. Conclusions By down-regulating miR-124, PF inhibits the proliferation and infammation of THP-1 cells, and promotes the apoptosis of THP-1 cells.

Effects of β Air Cooling and Subsequent Cold Rolling on Microstructure and Hardness of Zr702 Sheet

Haotian Guan,Lingguo Zeng,Zhijun Li,Linjiang Chai,Yufan Zhu,Yueyuan Wang,Qin Huang,Ke Chen,Liang‑yu Chen,Ning Guo 대한금속·재료학회 2021 METALS AND MATERIALS International Vol.27 No.2

In this work, a Zr702 sheet was subjected to β air cooling and then rolled to 15% reduction at room temperature, with theirdetailed microstructural characteristics characterized by electron channeling contrast imaging and electron backscatterdiffraction techniques. Results show that after the β air cooling, the prior equiaxed grains in the as-received material arecompletely transformed into Widmanstätten structures featured by coarse plates with typical phase transformation misorientationsbetween them. The subsequent 15% rolling allows both slip and twinning (especially the {10–12} type) to beactivated readily, leading to significant grain refinement and the appearance of misorientation angle peaks around 3°–5° and85°. Analyses on kernel average misorientations reveal that there exist very low residual strains in the β-air-cooled specimenwhile they are markedly increased after the 15% rolling and preferably distributed near low-angle and twin boundaries. Hardnessmeasurements show that the specimen hardness is evidently decreased from ~ 199 to ~ 170 HV after the β air cooling,which can be attributed to grain coarsening and the scattered orientations associated with the slow β → α transformation. For the 15%-rolled specimen, however, effective grain refinement by twinning and denser low-angle boundaries jointly leadto ~ 35% hardness increment to ~ 228 HV.

Peach Selection 97-41 with Double Flower

Ruijuan Ma,Mingliang Yu,Ping Du,Hongfeng Song,Zhijun Shen 한국원예학회 2006 한국원예학회 기타간행물 Vol.- No.-

Preparation of Cobalt Ferrite Nanoparticle-Decorated Boron Nitride Nanosheet Flame Retardant and Its Flame Retardancy in Epoxy Resin

Qiaoran Zhang,Zhiwei Li,Xiaohong Li,Laigui Yu,Zhijun Zhang,Zhishen Wu 성균관대학교(자연과학캠퍼스) 성균나노과학기술원 2019 NANO Vol.14 No.5

Boron nitride nanosheet (BNNS) decorated with cobalt ferrite nanoparticle (CFN) to afford CFN-BNNS nanohybrid was prepared via a simple hydrothermal route and was well characterized. Subsequently, the as-prepared CFN-BNNS nanohybrid was incorporated into epoxy resin (EP) with the introduction of a weak rotary magnetic field to achieve order orientation, in order to reduce the fire risk and toxic hazards using enhanced shielding effect of BNNS upon combustion. Findings demonstrate that the CFN-BNNS nanohybrid is composed of CFN nanoparticle uniformly dispersed on BNNS surface. Thermal analysis and cone calorimeter data show that the CFN-BNNS nanofiller among EP matrix contributes to improving the char residues and mechanical properties of EP and reducing its fire risk as well as toxic hazards, especially the ordered one is advantageous over the disordered one in reducing the fire risk and toxic hazard. This is because, on the one hand, the orderly aligned BNNS as the physical barrier can more effectively prevent the transfer and diffusion of oxygen and heat. On the other hand, CFN can catalyze the degradation of EP to afford excessive chars on polymer surface; and it is also liable to decomposition during combustion, thereby generating ferrite species to promote EP degradation as well as cobalt species to enhance the oxidation of CO.