检索结果-南通市图书馆

Development of EPDM composites reinforced by CNTs@SiO_(2)for thermal protection systems of aerospace propulsion:Significant improvement in oxidation and ablation resistance properties

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

Chinese Journal of Aeronautics 2024年第2期37卷 471-481页

作者： li wang Jiang li Kang li Yiwei wang Chenyang MA Science and Technology on Combustion Internal Flow and Thermal-Structure LaboratoryNorthwestern Polytechnical UniversityXi’an 710072China

Carbon Nanotubes(CNTs)reinforced Polymer-Matrix Composites(PMCs)is widely used as insulation materials in thermal protection system of aerospace ***,CNTs are prone to oxidation and have high thermal conductivities,which makes it difficult to improve the ablation resistance of insulation materials that contain ***_(2)was encapsulated onto the surface of CNTs(CNTs@SiO_(2)),which were then added to Ethylene Propylene Diene Monomer(EPDM)rubber to prepare the insulation *** analysis and ablation test were used to evaluate the resistance of the insulation materials to thermal oxidation and ***,scanning electron microscopy was performed to analyze their *** revealed that the addition of CNTs@SiO_(2)could visibly reduce the effects of hot corrosion and ablation on insulation *** C-CNTs@SiO_(2)-1 formulation had the best ablative ***,compared with the unencapsulated formulation(C-CNTs-10),the C-CNTs@SiO_(2)-1 formulation reduced the line ablation rate by 51%to 0.0130 mm/s after oxygen-acetylene ***,the ablation mechanism was investigated based on the effects of the CNTs@SiO_(2)additive on their ***,the improvement in ablation performance may be attributed to CNTs@SiO_(2)-induced decreases in thermal conductivity,improvement in the hot corrosion resistance in the char layer,and changes in the microstructure.

关键词： Polymer-Matrix Composites(PMCs) Carbon Nanotubes(CNTs) Thermal properties Ablation Oxidation

A multi-source information fusion layer counting method for penetration fuze based on TCN-LSTM

在线全文

学校读者我要写书评

暂无评论

Defence Technology（防务技术） 2024年第3期33卷 463-474页

作者： Yili wang Changsheng li Xiaofeng wang Nanjing University of Science and Technology Nanjing 210094JiangsuChina State Key Laboratory of Explosion Science and Technology Beijing Institute of TechnologyBeijing 100081China

When employing penetration ammunition to strike multi-story buildings,the detection methods using acceleration sensors suffer from signal aliasing,while magnetic detection methods are susceptible to interference from ferromagnetic materials,thereby posing challenges in accurately determining the number of *** address this issue,this research proposes a layer counting method for penetration fuze that incorporates multi-source information fusion,utilizing both the temporal convolutional network(TCN)and the long short-term memory(LSTM)recurrent *** leveraging the strengths of these two network structures,the method extracts temporal and high-dimensional features from the multi-source physical field during the penetration process,establishing a relationship between the multi-source physical field and the distance between the fuze and the target plate.A simulation model is developed to simulate the overload and magnetic field of a projectile penetrating multiple layers of target plates,capturing the multi-source physical field signals and their patterns during the penetration *** analysis reveals that the proposed multi-source fusion layer counting method reduces errors by 60% and 50% compared to single overload layer counting and single magnetic anomaly signal layer counting,*** model's predictive performance is evaluated under various operating conditions,including different ratios of added noise to random sample positions,penetration speeds,and spacing between target *** maximum errors in fuze penetration time predicted by the three modes are 0.08 ms,0.12 ms,and 0.16 ms,respectively,confirming the robustness of the proposed ***,the model's predictions indicate that the fitting degree for large interlayer spacings is superior to that for small interlayer spacings due to the influence of stress waves.

关键词： Penetration fuze Temporal convolutional network(TCN) Long short-term memory(LSTM) Layer counting Multi-source fusion

FlowGANAnomaly: Flow-Based Anomaly Network Intrusion Detection with Adversarial Learning

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2024年第1期33卷 58-71页

作者： Zeyi li Pan wang Zixuan wang School of Computer Science Nanjing University of Posts and Telecommunications School of Modern Posts Nanjing University of Posts and Telecommunications School of Internet of Things Nanjing University of Posts and Telecommunications

In recent years, low recall rates and high dependencies on data labelling have become the biggest obstacle to developing deep anomaly detection(DAD) techniques. Inspired by the success of generative adversarial networks(GANs) in detecting anomalies in computer vision and imaging, we propose an anomaly detection model called Flow GANAnomaly for detecting anomalous traffic in network intrusion detection systems(NIDS). Unlike traditional GAN-based approaches, which are composed of a flow encoder, a convolutional encoder-decoder-encoder, a flow decoder and a convolutional encoder, the architecture of this model consists of a generator(G) and a discriminator(D).Flow GANAnomaly maps the different types of traffic feature data from separate datasets to a uniform feature space,thus can capture the normality of network traffic data more accurately in an adversarial manner to mitigate the problem of the high dependence on data labeling. Moreover, instead of simply detecting the anomalies by the output of D, we proposed a new anomaly scoring method that integrates the deviation between the output of two Gs’ convolutional encoders with the output of D as weighted scores to improve the low recall rate of anomaly detection. We conducted several experiments comparing existing machine learning algorithms and existing deep learning methods(Auto Encoder and VAE) on four public datasets(NSL-KDD, CIC-IDS2017, CIC-DDo S2019, and UNSW-NB15). The evaluation results show that Flow GANAnomaly can significantly improve the performance of anomaly-based NIDS.

关键词： Deep learning Machine learning algorithms Computational modeling Network intrusion detection Imaging Telecommunication traffic Generative adversarial networks

Al/Hf ratio-dependent mechanisms of microstructure and mechanical property of nearly fully dense Al—Hf reactive material

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Defence Technology（防务技术） 2024年第1期31卷 230-241页

作者： Junbao li Weibing li Xiaoming wang ZNDY of Ministerial Key Laboratory Nanjing University of Science and TechnologyNanjing 210094China

This study proposed three types of Al—Hf reactive materials with particle size ratios(a),which were almost completely dense(porosity of<5.40%)owing to their preparation using hot-pressing *** characteristics and phas... 详细信息

This study proposed three types of Al—Hf reactive materials with particle size ratios(a),which were almost completely dense(porosity of40%)owing to their preparation using hot-pressing *** characteristics and phase composition were analyzed,and the influence of particle size ratios on dynamic mechanical behavior and damage mechanism were *** prepared sample with a=0.1 exhibited continuous wrapping of the Hf phase by the Al ***—Hf contact(continuous Hf phase)within the sample gradually increased with increasing a,and a small amount of fine Hf appeared for the sample with a=*** reactive materials exhibited clear strain-rate sensitivity,with flow stressσ0.05and failure strainεfincreasing approximately linearly with increasing strain rate.ε.It is found that the plastic deformation of the material increased with increasing strain *** a increased from 0.1 to 1,the flow stress gradually *** failure of the material was dominated by ductile fracture with a large Al phase plastic deformation band for lower a,while brittle fracture with crushed Hf particles occurred at higher ***,a constitutive model based on BP neural network was proposed to describe the stress-strain relationships of the materials,with an average relative error of 2.22%.

关键词： Reactive material Particle size Split Hopkinson pressure bar test Stressestrain relationship Impact failure BP neural network

Adaptive and augmented active anomaly detection on dynamic network traffic streams

在线全文

学校读者我要写书评

暂无评论

Frontiers of Information Technology & Electronic Engineering 2024年第3期25卷 446-460页

作者： Bin li Yijie wang li CHENG National Key Laboratory of Parallel and Distributed Computing College of ComputerNational University of Defense TechnologyChangsha 410073China College of System Engineering National University of Defense TechnologyChangsha 410073China

Active anomaly detection queries labels of sampled instances and uses them to incrementally update the detection model,and has been widely adopted in detecting network ***,existing methods cannot achieve desirable performance on dynamic network traffic streams because(1)their query strategies cannot sample informative instances to make the detection model adapt to the evolving stream and(2)their model updating relies on limited query instances only and fails to leverage the enormous unlabeled instances on *** address these issues,we propose an active tree based model,adaptive and augmented active prior-knowledge forest(A3PF),for anomaly detection on network trafic streams.A prior-knowledge forest is constructed using prior knowledge of network attacks to find feature subspaces that better distinguish network anomalies from normal *** one hand,to make the model adapt to the evolving stream,a novel adaptive query strategy is designed to sample informative instances from two aspects:the changes in dynamic data distribution and the uncertainty of *** the other hand,based on the similarity of instances in the neighborhood,we devise an augmented update method to generate pseudo labels for the unlabeled neighbors of query instances,which enables usage of the enormous unlabeled instances during model *** experiments on two benchmarks,CIC-IDS2017 and UNSW-NB15,demonstrate that A3PF achieves significant improvements over previous active methods in terms of the area under the receiver operating characteristic curve(AUC-ROC)(20.9%and 21.5%)and the area under the precision-recall curve(AUC-PR)(44.6%and 64.1%).

关键词： Active anomaly detection Network traffic streams Pseudo labels Prior knowledge of network attacks

Federated learning on non-IID and long-tailed data viadual-decoupling

在线全文

学校读者我要写书评

暂无评论

Frontiers of Information Technology & Electronic Engineering 2024年第5期25卷 728-741页

作者： Zhaohui wang Hongjiao li Jinguo li Renhao HU Baojin wang College of Computer Science and Technology Shanghai University of Electric PowerShanghai 201306China

Federated learning(FL),a cutting-edge distributed machine learning training paradigm,aims to generate a global model by collaborating on the training of client models without revealing local private *** co-occurrence of non-independent and identically distributed(non-IID)and long-tailed distribution in FL is one challenge that substantially degrades aggregate *** this paper,we present a corresponding solution called federated dual-decoupling via model and logit calibration(FedDDC)for non-IID and long-tailed *** model is characterized by three ***,we decouple the global model into the feature extractor and the classifier to fine-tune the components affected by the joint *** the biased feature extractor,we propose a client confidence re-weighting scheme to assist calibration,which assigns optimal weights to each *** the biased classifier,we apply the classifier re-balancing method for ***,we calibrate and integrate the client confidence re-weighted logits with the re-balanced logits to obtain the unbiased ***,we use decoupled knowledge distillation for the first time in the joint problem to enhance the accuracy of the global model by extracting the knowledge of the unbiased *** experiments demonstrate that on non-IID and long-tailed data in FL,our approach outperforms state-of-the-art methods.

关键词： Federated learning Non-IID Long-tailed data Decoupling learning Knowledge distillation

Insight into the adsorption behavior and mechanism of trace impurities from H_(2)O_(2) solution on functionalized zirconia by tuning the structure of amino groups

在线全文

学校读者我要写书评

暂无评论

Frontiers of Chemical Science and Engineering 2024年第5期18卷 83-94页

作者： Yu Meng Yitong wang Guozhu li Guozhu liu li wang Key Laboratory for Green Chemical Technology of Ministry Education School of Chemical Engineering and TechnologyTianjin UniversityTianjin 300072China Haihe Laboratory of Sustainable Chemical Transformations Tianjin 300192China

Primary,secondary and tertiary amino-functionalized zirconia(ZrO_(2)-NH_(2),ZrO_(2)-NH and ZrO_(2)-N)was synthesized by the postgrafting method for the adsorption removal of typical metallic ions,phosphate and total oxidizable carbon from a real H_(2)O_(2) ***_(2)-NH_(2),ZrO_(2)-NH and ZrO_(2)-N exhibited similar pore sizes and sequentially increased zeta *** adsorption results of single and binary simulated solutions showed that the removal efficiency increased in the order of Fe^(3+)>Al^(3+)>Ca^(2+)>Na^(+).There is competitive adsorption between metallic ions,and Fe^(3+) has an advantage over the other metals,with a removal efficiency of 90.7%.The coexisting phosphate could promote the adsorption of metallic ions,while total oxidizable carbon had no effect on *** adsorption results of the real H_(2)O_(2)solution showed that ZrO_(2)-NH_(2) exhibited the best adsorption affinity for metallic ions,as did phosphate and total oxidizable carbon,with a total adsorption capacity of 120.9 mg·g_(-1).Density functional theory calculations revealed that the adsorption process of metallic ions involves electron transfer from N atoms to metals and the formation of N-metal bonds.

关键词： adsorption metallic ion phosphate total oxidizable carbon zirconia H_(2)O_(2)

Semi-supervised remote sensing image scene classification with prototype-based consistency

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

Chinese Journal of Aeronautics 2024年第2期37卷 459-470页

作者： Yang li Zhang li Zi wang Kun wang Qifeng YU College of Aerospace Science and Engineering National University of Defense TechnologyChangsha 410073China

Deep learning significantly improves the accuracy of remote sensing image scene classification,benefiting from the large-scale ***,annotating the remote sensing images is time-consuming and even tough for *** neural networks trained using a few labeled samples usually generalize less to new unseen *** this paper,we propose a semi-supervised approach for remote sensing image scene classification based on the prototype-based consistency,by exploring massive unlabeled *** this end,we,first,propose a feature enhancement module to extract discriminative *** is achieved by focusing the model on the foreground ***,the prototype-based classifier is introduced to the framework,which is used to acquire consistent feature *** conduct a series of experiments on NWPU-RESISC45 and Aerial Image Dataset(AID).Our method improves the State-Of-The-Art(SOTA)method on NWPU-RESISC45 from 92.03%to 93.08%and on AID from 94.25%to 95.24%in terms of accuracy.

关键词： Semi-supervised learning Remote sensing Scene classification Prototype network Deep learning

UAV image target localization method based on outlier filter and frame buffer

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2024年第7期37卷 375-390页

作者： Yang wang Hongguang li Xinjun li Zhipeng wang Baochang ZHANG Institute of Unmanned System Beihang UniversityBeijing 100191China School of Electronics and Information Engineering Beihang UniversityBeijing 100191China Institute of Artificial Intelligence Beihang UniversityBeijing 100191China

With rapid development of UAV technology,research on UAV image analysis has gained *** the existing techniques of UAV target localization often rely on additional equipment,a method of UAV target localization based on depth estimation has been ***,the unique perspective of UAVs poses challenges such as the significant field of view variations and the presence of dynamic objects in the *** a result,the existing methods of depth estimation and scale recovery cannot be directly applied to UAV ***,there is a scarcity of depth estimation datasets tailored for UAV perspectives,which makes supervised algorithms *** address these issues,an outlier filter is introduced to enhance the applicability of depth estimation networks to target localization.A frame buffer method is proposed to achieve more accurate scale recovery,so as to handle complex scene textures in UAV *** proposed method demonstrates a 14.29%improvement over the *** with the average recovery results from UAV perspectives,the difference is only 0.88%,approaching the performance of scale recovery using ground truth ***,to overcome the limited availability of traditional UAV depth datasets,a method for generating depth labels from video sequences is *** to state-of-the-art methods,the proposed approach achieves higher accuracy in depth estimation and stands for the first attempt at target localization using image *** algorithm and dataset are available at https://***/uav-tan/uav-object-localization.

关键词： Object localization Deep learning Depth estimate Scale recovery Unmanned Aerial Vehicle(UAV)