检索结果-南通市图书馆

Journal of Thermal Science 2021年第2期30卷 493-503页

作者： CHEN Jiajia zhang liyong FU Yucan QIAN Ning JIANG Huafei YANG Lin College of Mechanical and Electronic Engineering Nanjing Forestry UniversityNanjing 210037China College of Mechanical and Electrical Engineering Nanjing University of Aeronautics and AstronauticsNanjing 210016China Jiangsu Wind Power Generation Engineering&Technology Center Nanjing 210023China

This study numerically analyzed the heat transfer characteristics outside the condenser of a rotating heat pipe grinding wheel(RHP-GW).The goal of this investigation is to determine the optimal structure and parameters for the condenser section of *** fin height(f=0-7 mm),rotational Reynolds number(Rer=1602-6408)and jet Reynolds number(Rej=42379-108302)were analyzed under input heat flux of 4000 W/m2.A fully developed flow was imposed at the outlet of the *** showed that the optimal heat transfer rate was obtained for fin height of 5 mm,which improved the average Nusselt number by 84%compared to the structure without fins.A critical Rej for each Rer that the impinging jet can reach the condenser section was *** critical Rej value increases with Rer,which is in the range from 42379 to 61215 and 61215 to 80050 for Rer=6408 and Rer=9610,respectively.

关键词： rotating heat pipe grinding wheel impinging jet heat transfer numerical analysis

A method for determination of deuterium impurity in the helium beam

在线全文

学校读者我要写书评

暂无评论

Science China(Physics,Mechanics & Astronomy) 2018年第5期61卷 53-56页

作者： Han Chen ShiWei Xu NingTao zhang Jun Hu KuoAng Li ShaoBo Ma XiChao Ruan XiaoDong Tang liyong zhang Institute of Modern Physics Chinese Academy of Sciences Department of Nuclear Physics China Institute of Atomic Energy Joint Department for Nuclear Physics Lanzhou University and Institute of Modern Physics Chinese Academy of Sciences National Astronomical Observatories of China Chinese Academy of Sciences

Both the LUNA（Laboratory for Underground Nuclear Astrophysics）collaboration in Europe and the JUNA（Jinping Underground Laboratory for Nuclear Astrophysics）collaboration in China are planning to study the key react... 详细信息

关键词：氢杂质天体物理学横梁氦实验室结束期原子合作

A hybrid-model optimization algorithm based on the Gaussian process and particle swarm optimization for mixed-variable CNN hyperparameter automatic search

在线全文

学校读者我要写书评

暂无评论

Frontiers of Information Technology & Electronic Engineering 2023年第11期24卷 1557-1573页

作者： Han YAN Chongquan ZHONG Yuhu WU liyong zhang Wei LU School of Control Science and Engineering Dalian University of TechnologyDalian 116024China

Convolutional neural networks(CNNs)have been developed quickly in many real-world ***,CNN’s performance depends heavily on its hyperparameters,while finding suitable hyperparameters for CNNs working in application fields is challenging for three reasons:(1)the problem of mixed-variable encoding for different types of hyperparameters in CNNs,(2)expensive computational costs in evaluating candidate hyperparameter configuration,and(3)the problem of ensuring convergence rates and model performance during hyperparameter *** overcome these problems and challenges,a hybrid-model optimization algorithm is proposed in this paper to search suitable hyperparameter configurations automatically based on the Gaussian process and particle swarm optimization(GPPSO)***,a new encoding method is designed to efficiently deal with the CNN hyperparameter mixed-variable ***,a hybrid-surrogate-assisted model is proposed to reduce the high cost of evaluating candidate hyperparameter ***,a novel activation function is suggested to improve the model performance and ensure the convergence *** experiments are performed on image-classification benchmark datasets to demonstrate the superior performance of GPPSO over state-of-the-art ***,a case study on metal fracture diagnosis is carried out to evaluate the GPPSO algorithm performance in practical *** results demonstrate the effectiveness and efficiency of GPPSO,achieving accuracy of 95.26%and 76.36%only through 0.04 and 1.70 GPU days on the CIFAR-10 and CIFAR-100 datasets,respectively.

关键词： Convolutional neural network Gaussian process Hybrid model Hyperparameter optimization Mixed-variable Particle swarm optimization

A proposed direct measurement of cross section at Gamow window for key reaction 19F(p,α) 16O in Asymptotic Giant Branch stars with a planned accelerator in CJPL

在线全文

学校读者我要写书评

暂无评论

Science China(Physics,Mechanics & Astronomy) 2016年第5期59卷 76-82页

作者： JianJun He ShiWei Xu ShaoBo Ma Jun Hu liyong zhang ChangBo Fu NingTao zhang Gang Lian Jun Su YunJu Li ShengQuan Yan YangPing Shen SuQing Hou BaoLu Jia Tao zhang XiaoPeng zhang Bing Guo Shigeru Kubono WeiPing Liu Institute of Modern Physics Chinese Academy of Sciences Lanzhou 730000 China University of Chinese Academy of Sciences Beijing 100049 China Department of Physics Shanghai Jiao Tong University Shanghai 200240 China China Institute of Atomic Energy P. O. Box 275(10) Beijing 102413 China RIKEN Nishina Center 2-1 Hirosawa Wako Saitama 351-0198 Japan

In 2014, the National Natural Science Foundation of China （NSFC） approved the Jinping Underground Nuclear Astrophysics laboratory （JUNA） project, which aims at direct cross-section measurements of four key stellar nuclear reactions right down to the Gamow windows. In order to solve the observed fluorine overabundances in Asymptotic Giant Branch （AGB） stars, measuring the key 19F（p,a）16O reaction at effective burning energies （i.e., at Gamow window） is established as one of the scientific research sub-projects. The present paper describes this sub-project in details, including motivation, status, experimental setup, yield and background estimation, aboveground test, as well as other relevant reactions.

关键词： Jinping Underground Nuclear Astrophysics laboratory (JUNA) direct measurement Gamow window cross section AGBstar nucleosynthesis

Filter Pruning Based on Coordination of Layers for Deep Neural Networks

在线全文

学校读者我要写书评

暂无评论

Filter Pruning Based on Coordination of Layers for Deep Neur...

Incomplete Data Imputation Based on Single Output Sub-network Array and Missing Value Variable Description

第43届中国控制会议

作者： Xing Luan liyong zhang School of Control Science and Engineering Dalian University of Technology

Convolutional neural networks have achieved remarkable success in the field of computer vision. However,deploying them to edge devices is challenging due to their numerous parameters and computational complexity. In this paper,we propose a new network pruning method called Filter Pruning based on Coordination of Layers(FPCL), which reduces the size of the network by pruning some relatively unimportant filters while ensuring low accuracy loss. In order to maintain the coordination of the feature extraction by filters in the adjacent network layers, we take the influence of the filter on the loss function during the images of dataset passing through the network as the attribute of the filter, cluster filters in the adjacent layers, and then prune the filters in the cluster. During pruning, we ensure that filters from different layers within the same cluster are properly preserved. To avoid the loss of model generalization after pruning, we fine-tune the pruned model and then re-evaluate all filters, including the pruned ones, to prune again. The process of pruning and fine-tuning is iterated multiple times, with an increasing number of filters pruned each time, until the desired number of pruned filters is reached. To validate the effectiveness of the FPCL method, we pruned various models on multiple datasets. Compared to other advanced methods,the pruned model by FPCL achieves higher testing accuracy with fewer parameters.

关键词： Filter pruning Clustering Coordination Convolutional neural network Image classification

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

Incomplete Data Imputation Based on Single Output Sub-networ...

Hypoxic/ischemic preconditioning attenuate PKCδ-medi-ated injury in patients and mice with cerebral infarction

第43届中国控制会议

作者： Meitong Chen liyong zhang School of Control Science and Engineering Dalian University of Technology

Incomplete data widely exists in the real world, and it is of great significance to reasonably impute the missing values of incomplete data. In this paper, we construct a single output sub-network array of incomplete data, which can mine the association relationships between each dimensional attribute of incomplete data and other attributes. At the same time, to address the problem of incomplete input of the model caused by the presence of missing values, a new training scheme of the model is designed that treats missing values as variables and allows them to participate in network training together with existing values,thereby maximizing the use of known information. During the training process, network parameters and missing values are updated collaboratively, and the parameters between single output sub-networks are adjusted to each other by missing value variables. Finally, the training of the model and the imputation of the missing values are completed simultaneously. The experiment results on several public datasets demonstrate the effectiveness of the proposed method.

关键词： Incomplete data Imputation Single output sub-network array Missing value variables Collaborative train

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

Journal of Translational Neuroscience 2021年第3期6卷 19-29页

作者： Weiwei Yang Shengli Xu liyong zhang Zidong Wang Department of Neurobiology Xuanwu Hospital of Capital Medical UniversityBeijing 100053China National Clinical Research Center for Geriatric Disorders Beijing 100053China Department of Neurosurgery Liaocheng People’s HospitalLiaocheng 252000China

Objective:cerebral ischemic/hypox-ic preconditioning(I/HPC)is an endogenous strategy in which brief periods of sublethal ischemia/hypoxia render neural tissues resistant to subsequent ischemic/hypoxic *** phenomenon has been found in the brain,heart,liver,intestine,muscle,kidneys,and ***-ever,whether HPC has a protective effect on secondary cerebral ischemic injury or protein kinase Cδ(PKCδ)within ischemic patients and animal models is still ***:using a hypoxic preconditioned mouse model and a middle cerebral artery occlusion mouse mod-el,combined with 2,3,5-triphenyl tetrazolium chloride(TTC)staining,SDS-polyacrylamide gel electrophoresis(SDS-PAGE),and Western blot,we observed changes in infarction size,density,edema ratio,and changes in PKCδand membrane translocation within the ischemic cortex of the middle cerebral artery occlusion(MCAO)***:HPC can attenuate neurological deficits and cerebral ischemic injuries of mice following MCAO,including decreases in infarct size,edema ratio,densities of infarct area,and neuron *** addition,HPC inhib-its PKCδmembrane translocation in the penumbra of the MCAO-induced ischemic *** found that admin-istration of PKCδ-specific inhibitor dV1-1 mimics the neuroprotective effects of HPC,and nonisoform-specif-ic activation of PKC can partially abolish HPC-induced *** preconditioning decreased the levels of PKCδin the serum of patients with cerebral in-farction and reduced the cerebral nerve damage caused by ***:hypoxic/ischemic precondi-tioning attenuates PKCδ-mediated injury in patients and *** findings enrich our understanding of the sig-nal transduction mechanism underlying cerebral HPC and provide clues to developing medicine against ischemia/hypoxia-induced cerebral injuries.

关键词： hypoxic preconditioning(HPC) middle cerebral artery occlusion(MCAO) protein kinase C(PKC) penumbra neuroprotection

Image Classification Using Convolutional Neural Network Based on Feature Selection for Edge Computing

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

Image Classification Using Convolutional Neural Network Base...

基于最优分组原则的多传感器分组加权融合算法

第40届中国控制会议

作者： Pingchang Hao liyong zhang School of Control Science and Engineering Dalian University of Technology

In the field of image classification,convolutional neural networks are widely used because of its efficient feature extraction ***,some of the features extracted by convolutional neural networks have insignificant difference among classes,which not only have little contribution to classification,but also cause the increasing of complexity of *** this will cause some difficulties in embedded deployment and application in many fields such as edge *** selection methods have the ability to select features that are helpful to classification,reduce the dimensions of features and improve the classification *** view of the existence of labels in training of image classification,and considering the potential role of labels information in feature selection,we propose an image classification method using convolutional neural network based on fuzzy clustering to select *** the basis of image features extracted by convolution neural network,this method reconstructs the classifier according to the feature selection algorithm by means of fuzzy clustering,so as to select the valuable features to participate in the image ***,we divide the features extracted by convolutional neural network from the image of training sets according to labels,and use Fuzzy C-Means to obtain the distribution of each feature on each *** we design a measure of the distribution difference between classes according to the cluster centers and membership degree matrices,to select the features with obvious inter-class difference to participate in *** experimental results on several image datasets verify the effectiveness of the proposed method.

关键词： Image feature selection Image classification Distribution differences between classes Convolutional neural network Fuzzy C-Means

来源： cnki会议评论

在线全文

cnki会议

学校读者我要写书评

暂无评论

仪器仪表学报 2008年第1期29卷 200-205页

作者：张立勇张利李丹大连理工大学电子与信息工程学院大连116024

多传感器对某一状态进行测量时,将多传感器分组后进行加权融合处理,可实现对待测状态较为准确的估计,但多传感器的组合方式不同,估计效果的差异很大。本文针对分组融合算法中如何选择传感器的分组组合方式进行了研究,依据分组中传感器... 详细信息

多传感器对某一状态进行测量时,将多传感器分组后进行加权融合处理,可实现对待测状态较为准确的估计,但多传感器的组合方式不同,估计效果的差异很大。本文针对分组融合算法中如何选择传感器的分组组合方式进行了研究,依据分组中传感器的个数和各组间传感器测量方差的大小关系,提出了使融合估计值最优的分组原则。并将这些原则加以数学抽象,从理论上证明了基于最优分组原则的分组融合算法的最小方差特性。最优分组原则为评价和选择传感器的分组组合方式提供了一个理论依据。实例分析表明:该分组原则极大地缩小了寻找最优分组组合方式的范围,有效地解决了在如何分组问题上的盲目性,从而使得多传感器测量时的在线优化分组得以实现。

关键词：多传感器分组加权融合分组组合方式最优分组原则最小方差

Research of Protected Cultivation Techniques of Toona sinensis on Mountain Slopes

在线全文

学校读者我要写书评

暂无评论

Asian Agricultural Research 2018年第3期10卷 78-82页

作者： Qingzhen YIN Chengtong GU liyong zhang Mingyue GU Institute of Cash Crops of Hebei Academy of Agriculture and Forestry Sciences Shijiazhuang 050051 China Luquan Gujia Toona sinensis Specialized Cooperative Shijiazhuang 050200 China Hebei Chemical and Pharmaceutical College Shijiazhuang 050026 China

In order to prolong the supply period and increase the economic value of Toona sinensis,the protected cultivation of T. sinensis was carried out on mountainous slope to achieve stagger-season planting and the year round supply. The results showed that in late October to midNovember,the temperature was moderate( 1-10℃) and suitable for the lifting of seedlings and temporary planting dormancy. The line spacing of rooted seedlings in greenhouse was( 22-24) cm ×( 4-5) cm and planting density was 900 000-1 200 000 seedlings/ha. The line spacing of seedlings was( 14-13) cm ×( 3-4) cm and planting density was 1 950 000-2 100 000 seedlings/ha. T. sinensis began to sprout at 8-10℃ and stretch at 10-12℃. When the temperature was stable at 14℃ or above,T. sinensis started bolting and emerging leaves. The terminal buds sprouted in 40-50 d for the T. sinensis planted in greenhouse with the day temperature of 15-22℃ and night temperature over 10℃; the terminal buds sprouted in about 15 d for those planted in the greenhouse with day temperature of 20-25℃ and night temperature over 10℃. T. sinensis could be picked 3 times when planted in greenhouse,and the yield was about 24 720 kg/ha.

关键词： Toona sinensis Protected cultivation Temporary planting Dormancy Cultivation techniques