检索结果-南通市图书馆

郑州大学学报（工学版） 2024年第2期45卷 12-19页

作者：张震王晓杰晋志华马继骏郑州大学电气与信息工程学院河南郑州450001 河南省交通调度指挥中心河南郑州450001

为了提高道路交通标志的检测速度,提出一种基于轻量化YOLOv5的改进模型。首先,使用Ghost卷积和深度分离卷积(DWConv)构建新的主干模块,减少计算量和参数量;引入加权特征融合网络(BiFPN)结构,增强特征融合能力;将CIoU损失函数替换为SIoU... 详细信息

为了提高道路交通标志的检测速度,提出一种基于轻量化YOLOv5的改进模型。首先,使用Ghost卷积和深度分离卷积(DWConv)构建新的主干模块,减少计算量和参数量;引入加权特征融合网络(BiFPN)结构,增强特征融合能力;将CIoU损失函数替换为SIoU损失函数,关注真实锚框与预测的角度信息,提升检测精度。其次,对TT100K数据集进行优化,筛选出标签个数大于200的交通标志图片和标注信息共24类。最后,实验结果取得84%的准确率、81.2%的召回率和85.4%的所有类别平均精确率的平均值mAP@0.5,相比原始YOLOv5,参数量减少29.0%,计算量减少29.4%,mAP@0.5仅下降0.1百分点,检测帧率提升了34帧/s。使用改进后的模型进行检测,检测速度有了明显提升,基本达到了在保持检测精度的基础上压缩模型的目的。

关键词：交通标志检测轻量化YOLOv5 SIoU损失函数 Ghost卷积 TT100K BiFPN

维普期刊数据库博看期刊

STGSA:A Novel Spatial-Temporal Graph Synchronous Aggregation Model for Traffic Prediction

在线全文

学校读者我要写书评

暂无评论

IEEE/CAA Journal of Automatica Sinica 2023年第1期10卷 226-238页

作者： Zebing Wei Hongxia Zhao Zhishuai Li xiaojie Bu Yuanyuan Chen Xiqiao zhang Yisheng Lv Fei-Yue wang the State Key Laboratory for Management and Control of Complex Systems Institute of AutomationChinese Academy of SciencesBeijing 100190China the school of Artificial Intelligence University of Chinese Academy of SciencesBeijing 100049China the Shandong Key Laboratory of Smart Transportation(Preparation) Shandong Jiao Tong UniversityShanghai 200240China the College of Science China University of PetroleumQingdao 266580China the Department of Transportation engineering School of Transportation Science and TechnologyHarbin Institute of TechnologyHarbin 150090China IEEE

The success of intelligent transportation systems relies heavily on accurate traffic prediction,in which how to model the underlying spatial-temporal information from traffic data has come under the *** existing frameworks typically utilize separate modules for spatial and temporal correlations ***,this stepwise pattern may limit the effectiveness and efficiency in spatial-temporal feature extraction and cause the overlook of important information in some ***,it is lacking sufficient guidance from prior information while modeling based on a given spatial adjacency graph(e.g.,deriving from the geodesic distance or approximate connectivity),and may not reflect the actual interaction between *** overcome those limitations,our paper proposes a spatial-temporal graph synchronous aggregation(STGSA)model to extract the localized and long-term spatial-temporal dependencies ***,a tailored graph aggregation method in the vertex domain is designed to extract spatial and temporal features in one graph convolution *** each STGSA block,we devise a directed temporal correlation graph to represent the localized and long-term dependencies between nodes,and the potential temporal dependence is further fine-tuned by an adaptive weighting ***,we construct an elaborated spatial adjacency matrix to represent the road sensor graph by considering both physical distance and node similarity in a datadriven ***,inspired by the multi-head attention mechanism which can jointly emphasize information from different r epresentation subspaces,we construct a multi-stream module based on the STGSA blocks to capture global *** projects the embedding input repeatedly with multiple different ***,the predicted values are generated by stacking several multi-stream *** experiments are constructed on six real-world datasets,and numerical results show that the proposed STGSA

关键词： Deep learning graph neural network(GNN) multistream spatial-temporal feature extraction temporal graph traffic prediction

Analysis and Design of Four-Plate Capacitive Wireless Power Transfer System for Undersea Applications

在线全文

学校读者我要写书评

暂无评论

CES Transactions on Electrical Machines and Systems 2021年第3期5卷 202-211页

作者： Lei Yang Yuanqi zhang xiaojie Li Jiale Jian Zhe wang Jingjing Huang Li Ma Xiangqian Tong school of Electrical Enginerring Xi’an Technology of UniversityXi’an710048China the school of Electronic and information engineering Xi’an Jiaotong UniversityXi’an 710049China

This paper presents a four-plate undersea capacitive wireless power transfer(CPT)system for underwater applications such as autonomous underwater vehicles(AUVs).Generally,a CPT system transfers the power based on electric *** complex resonant compensation networks are used to make the CPT system work in the resonant *** resonant voltage is always very *** will be a big challenge to the human *** this paper,a virtual electrons periodic reciprocating flow theory is proposed for the CPT *** one switching cycle,the electrons firstly flow in the forward direction through the forward path and then flow in the inverse direction through the inverse *** CPT system has been deeply studied with the vacuum dielectric or the air ***,for the CPT system,there are few papers to show the underwater *** this paper,an undersea four-plate CPT system is designed and studied in the underwater *** two coupling capacitors and other elements of the CPT system could build a closed-loop path.A small value inductor is adapted as a resonant compensation network for the four-plate CPT *** DC voltage is inverted to the AC voltage in the primary side with the single-phase full-bridge *** resonant voltage is rectified to the DC voltage in the secondary side with the single-phase full-bridge diode rectifier.A 100 W power level CPT system is constructed to verify the theory analysis and the *** theory analysis is verified by the simulated and experimental *** stable output voltage and load power are achieved in this paper.

关键词： Wireless power transfer system capacitive underwater applications autonomous underwater vehicles(AUVs)

Monitoring of Continuous Steel Casting Process Based on Independent Component Analysis

在线全文

学校读者我要写书评

暂无评论

Monitoring of Continuous Steel Casting Process Based on Inde...

COVID-19疫情防控中的中国公众舆情时空演变特征

2010 Chinese Control and Decision Conference

作者： Zhenping Ji~1,xiaojie zhang~1,Canrong wang~2 1.school of information Science and engineering,Shenyang Ligong University,Shengyang,110159,China 2.Sanming Iron and Steel Works,Sanming,365000,China.

Multivariate statistical process control(MSPC)has been successfully applied to performance monitoring and fault diagnosis for industrial ***,traditional MSPC are based upon the assumption that the separated latent variables must be subject to normal probability distribution,which can not be satisfied for continuous steel casting *** this paper,independent component analysis(ICA)is introduced to model non-Gaussian data from continuous steel casting process and improve the monitoring performance of process,which can overcome the need of the data *** basic idea of our approach is to use ICA to extract the essential independent components that drive a process and to combine them with process monitoring techniques.I,I and SPE charts are proposed as monitoring *** application results show the advantages of ICA monitoring in comparison to principal component analysis(PCA)monitoring.

关键词： Continuous Steel Casting PCA Independent Component Analysis(ICA) Process Monitoring

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

地理学报 2020年第11期75卷 2490-2504页

作者：王卷乐张敏韩雪华王晓洁郑莉中国科学院地理科学与资源研究所资源与环境信息系统国家重点实验室北京100101 中国科学院大学北京100049 山东理工大学建筑工程学院淄博255000 防灾科技学院三河065201 江苏省地理信息资源开发与利用协同创新中心南京210023

COVID-19疫情是全球面临的重大公共卫生危机。客观认识疫情期间的公众舆情响应和区域差异,对于提高重大公共卫生事件的政策调控和科学治理具有现实意义。本文以新浪微博为数据源,基于潜在狄利克雷分配主题模型和随机森林算法构建了主题... 详细信息

COVID-19疫情是全球面临的重大公共卫生危机。客观认识疫情期间的公众舆情响应和区域差异,对于提高重大公共卫生事件的政策调控和科学治理具有现实意义。本文以新浪微博为数据源,基于潜在狄利克雷分配主题模型和随机森林算法构建了主题抽取和分类模型,识别微博文本中的13个舆情话题,并从数量、空间、时间、内容等方面分析了2020年1月9日—3月10日在湖北省、京津冀、长三角、珠三角、成渝等城市群及沿边口岸等重点区域分布特点。结果表明:中国公众的响应总体是理性和积极的,但各舆情话题在区域内部的空间分布差异明显。各区域热点分布中,京津冀以首都北京为中心,长三角以上海为中心,辅以南京、杭州等热点,珠三角以广州、深圳为两核,湖北省以武汉为中心。建议应持续加强重点区域的疫情舆情关注和因地制宜的差异化精准响应。

关键词： COVID-19 疫情防控公众舆情区域差异

基于多智能体元强化学习的车联网协同服务缓存和计算卸载

在线全文

学校读者我要写书评

暂无评论

通信学报 2021年第6期42卷 118-130页

作者：宁兆龙张凯源王小洁郭磊重庆邮电大学通信与信息工程学院重庆400065 大连理工大学软件学院辽宁大连116620

为了降低求解优化问题的难度,提出一种双层的多路侧单元(RSU)协同缓存框架将问题进行解耦。外层采用多智能体元强化学习方法,在每个本地智能体进行决策学习的同时,采用长短期记忆网络作为元智能体来平衡本地决策并加速学习过程,从而得... 详细信息

为了降低求解优化问题的难度,提出一种双层的多路侧单元(RSU)协同缓存框架将问题进行解耦。外层采用多智能体元强化学习方法,在每个本地智能体进行决策学习的同时,采用长短期记忆网络作为元智能体来平衡本地决策并加速学习过程,从而得到最优的RSU缓存策略;内层采用拉格朗日乘子法求解最佳协同卸载策略,实现RSU间的任务分配。基于杭州真实交通数据的实验表明,所提方法具有理想的能效性能,并且能够在大规模任务流下保持网络稳健性。

关键词：车联网边缘服务缓存协同卸载元强化学习

在线全文

学校读者我要写书评

暂无评论

基于改进密度峰值聚类的异常流量检测

济南大学学报（自然科学版） 2022年第5期36卷 597-602页

作者：任艳徐春张蕾汪晓洁新疆财经大学信息管理学院新疆乌鲁木齐830012 新疆大学信息科学与工程学院新疆乌鲁木齐830046

针对网络异常流量检测技术准确率较低、簇的误划分等问题,提出基于改进密度峰值聚类算法的网络异常流量检测方案;首先对网络流量数据进行预处理和分组乱序,然后计算相应属性值并利用局部密度发现簇中心点,最后采用一种新的标签传递方式... 详细信息

针对网络异常流量检测技术准确率较低、簇的误划分等问题,提出基于改进密度峰值聚类算法的网络异常流量检测方案;首先对网络流量数据进行预处理和分组乱序,然后计算相应属性值并利用局部密度发现簇中心点,最后采用一种新的标签传递方式形成相应的簇群直至处理完所有数据。结果表明,相对于k均值算法和具有噪声的基于密度的聚类算法,基于改进的密度峰值聚类算法提升了网络异常流量的检测准确率,综合性能较优。

关键词：网络安全异常流量检测聚类密度峰值局部密度

AIE活性抗冻水凝胶可实现低温环境下多级信息的加密与解密

在线全文

学校读者我要写书评

暂无评论

engineering 2023年第4期23卷 82-89,M0005页

作者：睢晓洁汪晓东蔡诚诚马君怡杨静张雷 Department of Biochemical engineering School of Chemical Engineering and TechnologyFrontier Science Center for Synthetic Biology and Key Laboratory of Systems Bioengineering(MOE)Tianjin UniversityTianjin 300350China

抗冻水凝胶可在低温环境下调控内部水分子的冻结状态,进而维持其优异特性(如智能响应性、离子运输能力等)以及拓展其在低温环境下的应用。本文开发了一系列具有AIE活性的抗冻水凝胶,其可用于低温环境下的信息加密/解密。根据水凝胶中甜... 详细信息

抗冻水凝胶可在低温环境下调控内部水分子的冻结状态,进而维持其优异特性(如智能响应性、离子运输能力等)以及拓展其在低温环境下的应用。本文开发了一系列具有AIE活性的抗冻水凝胶,其可用于低温环境下的信息加密/解密。根据水凝胶中甜菜碱浓度的不同使其具有可调的冻结温度(T_(f))。当温度高于/低于Tf时,编码于水凝胶内的AIE分子表现出不发光/发光行为,从而实现信息的加密/解密。此外,通过调控水凝胶的冷冻程序,或者在水凝胶中原位引入具有光热效应的硫化铜纳米粒子并结合一定的照射条件,可实现多级信息的读取,从而增强信息的安全性。最后,由于该水凝胶对外界温度波动表现出不可逆性,因此可作为无需外界供能的超低温防伪标签,用于实时可视化监测冷链运输过程中细胞的活性。

关键词： Freeze-tolerant hydrogels Aggregation-induced emission Encryption Decryption Anticounterfeiting

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

基于增强现实的机械类专业情景教学探索

无线互联科技 2020年第10期17卷 99-100页

作者：朱继华王仁宗张素香中原工学院机电学院河南郑州450007

疫情防控期间在线教学的重要性日益提升。随着信息技术高速发展,虚拟现实技术与增强现实技术对促进学生的知识建构同样意义重大。文章通过对两种技术的分析比较,总结出增强现实技术在课堂教学中的推广优势,为其进入机械类专业情景教学... 详细信息

疫情防控期间在线教学的重要性日益提升。随着信息技术高速发展,虚拟现实技术与增强现实技术对促进学生的知识建构同样意义重大。文章通过对两种技术的分析比较,总结出增强现实技术在课堂教学中的推广优势,为其进入机械类专业情景教学建设领域提供借鉴。

关键词：虚拟现实增强现实情景教学增强现实应用