检索结果-南通市图书馆

Univariate imputation method for recovering missing data in wastewater treatment process

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Chemical Engineering 2023年第1期53卷 201-210页

作者： Honggui Han meiting sun Huayun Han Xiaolong Wu Junfei Qiao Faculty of Information Technology Beijing University of TechnologyBeijing 100124China Beijing Key Laboratory of Computational Intelligence and Intelligent System Beijing 100124China Engineering Research Center of Digital Community Ministry of EducationBeijing 100124China

High-quality data play a paramount role in monitoring,control,and prediction of wastewater treatment process(WWTP)and can effectively ensure the efficient and stable operation of *** values seriously degrade the accuracy,reliability and completeness of the data quality due to network collapses,connection errors and data transformation *** these cases,it is infeasible to recover missing data depending on the correlation with other *** tackle this issue,a univariate imputation method(UIM)is proposed for WWTP integrating decomposition method and imputation ***,the seasonal-trend decomposition based on loess method is utilized to decompose the original time series into the seasonal,trend and remainder components to deal with the nonstationary characteristics of WWTP ***,the support vector regression is used to approximate the nonlinearity of the trend and remainder components respectively to provide estimates of its missing values.A self-similarity decomposition is conducted to fill the seasonal component based on its periodic ***,all the imputed results are merged to obtain the imputation ***,six time series of WWTP are used to evaluate the imputation performance of the proposed UIM by comparing with existing seven methods based on two *** experimental results illustrate that the proposed UIM is effective for WWTP time series under different missing ***,the proposed UIM is a promising method to impute WWTP time series.

关键词： Univariate Self-similarity Waste water Algorithm Integration

Finite Horizon H2/H∞ Control for Discrete-Time Time-Delay Stochastic Systems with Infinite Markov Jumps

在线全文

学校读者我要写书评

暂无评论

Finite Horizon H2/H∞ Control for Discrete-Time Time-Delay S...

基于Transformer方法的任意风格迁移策略

第37届中国控制会议

作者： meiting sun Ting Hou College of Mathematics and Systems Science Shandong University of Science and Technology

In this paper,we mainly discuss finite horizon H2/H∞ control for discrete-time time-delay systems with infinite Markov ***,we present a stochastic bounded real lemma,which is crucial to solve mixed H2/H∞ control problem of the considered *** the variable substituting method,we transform a discrete system with control delay into a standard non-delay *** a sufficient and necessary condition is presented for the H2/H∞ control problem of infinite Markov jump systems.

关键词： H2/H∞ Control Finite Horizon Infinite Markov Jumps Time-Delay Systems

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

中国图象图形学报 2023年第11期28卷 3536-3549页

作者：孙梅婷代龙泉唐金辉南京理工大学计算机科学与工程学院南京210094

目的任意风格迁移是图像处理任务的重要分支,卷积神经网络作为其常用的网络架构,能够协助内容和风格信息的提取与分离,但是受限于卷积操作感受野,只能捕获图像局部关联先验知识;而自然语言处理领域的Transformer网络能够突破距离限制,... 详细信息

目的任意风格迁移是图像处理任务的重要分支,卷积神经网络作为其常用的网络架构,能够协助内容和风格信息的提取与分离,但是受限于卷积操作感受野,只能捕获图像局部关联先验知识;而自然语言处理领域的Transformer网络能够突破距离限制,捕获长距离依赖关系,更好地建模全局信息,但是因为需要学习所有元素间的关联性,其表达能力的提高也带来了计算成本的增加。鉴于风格迁移过程与句子翻译过程的相似性,提出了一种混合网络模型,综合利用卷积神经网络和Transformer网络的优点并抑制其不足。方法首先使用卷积神经网络提取图像高级特征,同时降低图像尺寸。随后将提取的特征送入Transformer中,求取内容特征与风格特征间的关联性,并将内容特征替换为风格特征的加权和,实现风格转换。最后使用卷积神经网络将处理好的特征映射回图像域,生成艺术化图像。结果与5种先进的任意风格迁移方法进行定性和定量比较。在定性方面,进行用户调查,比较各方法生成图像的风格化效果,结果表明本文网络生成的风格化图像渲染效果更受用户喜爱;在定量方面,比较各方法的风格化处理速度,结果表明本文网络风格化速率排名第3,属于可接受范围内。此外,本文与现有的基于Transformer的任意风格迁移方法进行比较,突出二者间差异;对判别网络进行消融实验,表明判别网络的引入能够有效提升图像的光滑度和整洁度;最后,将本文网络应用于多种风格迁移任务,表明本文网络具有灵活性。结论本文提出的混合网络模型,综合了卷积神经网络和Transformer网络的优点,同时引入了判别网络,使生成的风格化图像更加真实和生动。

关键词：计算机视觉图像处理任意风格迁移注意力机制 Transformer

基于贝叶斯改进结构算法的回转窑故障诊断模型研究

在线全文

学校读者我要写书评

暂无评论

中国机械工程 2017年第18期28卷 2143-2151页

作者：刘彬刘永记刘浩然李雷孙美婷燕山大学信息科学与工程学院秦皇岛066004 燕山大学电气工程学院秦皇岛066004

针对现有改进互信息爬山(MI&HC)算法精度低、耗时长及简化爬山(SHC)算法产生大量冗余边的问题,提出一种新的结构学习算法,即改进爬山(IHC)算法。通过计算互信息链得到贝叶斯初始结构,利用条件独立性测试以及对孤立节点进行处理来加边补... 详细信息

针对现有改进互信息爬山(MI&HC)算法精度低、耗时长及简化爬山(SHC)算法产生大量冗余边的问题,提出一种新的结构学习算法,即改进爬山(IHC)算法。通过计算互信息链得到贝叶斯初始结构,利用条件独立性测试以及对孤立节点进行处理来加边补充贝叶斯初始结构得到完全结构,利用改进的爬山搜索算子对完全结构进行搜索直到得出最优结构。将该算法与爬山(HC)算法、MI&HC算法、SHC算法进行比较,仿真结果表明,IHC算法能够得到较高准确率的模型,时间开销最小而且产生的冗余边数远远少于SHC算法产生的冗余边数。最后基于IHC算法,结合某回转窑数据进行训练,得到了回转窑工艺参数的故障诊断模型,对回转窑的烧成带温度实现了较为准确的故障诊断。

关键词：改进贝叶斯结构算法互信息及条件独立性测试故障诊断水泥回转窑

在线全文

学校读者我要写书评

暂无评论

基于蚁群节点寻优的贝叶斯网络结构算法研究

仪器仪表学报 2017年第1期38卷 143-150页

作者：刘浩然孙美婷李雷刘永记刘彬燕山大学信息科学与工程学院秦皇岛066004 燕山大学电气工程学院秦皇岛066004

K2算法是学习贝叶斯网络结构的经典算法。针对K2算法依赖最大父节点数和节点序的不足,以及蚁群算法搜索空间庞大的问题,提出了一种新的贝叶斯结构学习算法-MWST-ACO-K2算法。该算法通过计算互信息建立最大支撑树(MWST),得到最大父节点数... 详细信息

K2算法是学习贝叶斯网络结构的经典算法。针对K2算法依赖最大父节点数和节点序的不足,以及蚁群算法搜索空间庞大的问题,提出了一种新的贝叶斯结构学习算法-MWST-ACO-K2算法。该算法通过计算互信息建立最大支撑树(MWST),得到最大父节点数;然后利用蚁群算法(ACO)搜索最大支撑树,获得节点顺序;最后结合K2算法得到最优的贝叶斯网络结构。仿真实验结果表明,该方法不仅解决了K2算法依赖先验知识的问题,而且减少了蚁群算法的搜索空间,简化了搜索机制,得到较好的贝叶斯结构。最后将该算法应用到冀东水泥回转窑的实际数据中,构建水泥回转窑的贝叶斯网络结构,提高了故障诊断的准确率。

关键词：互信息蚁群优化 K2算法贝叶斯网络结构学习水泥回转窑

基于知识图谱的我国数字档案馆研究热点及趋势探析

在线全文

学校读者我要写书评

暂无评论

兰台世界 2022年第6期 54-58页

作者：孙美婷黑龙江大学信息管理学院

随着信息技术的推动,有关数字档案馆的研究越来越被学者重视。为进一步促进其研究与发展,本文采用CiteSpace可视化分析软件,从文献的发文数量、发文作者及研究机构等方面,对CNKI所收录的有关数字档案馆的相关文献进行统计,旨在探究我国... 详细信息

随着信息技术的推动,有关数字档案馆的研究越来越被学者重视。为进一步促进其研究与发展,本文采用CiteSpace可视化分析软件,从文献的发文数量、发文作者及研究机构等方面,对CNKI所收录的有关数字档案馆的相关文献进行统计,旨在探究我国数字档案馆领域的研究热点及未来的研究发展方向,以期为后续开展相关主题研究提供参考和借鉴。

关键词：数字档案馆 CiteSpace 可视化分析

Divergent responses of runoff to climate change in the upper basins of the Third Pole dominated by westerlies and monsoon

在线全文

学校读者我要写书评

暂无评论

Science China Earth Sciences 2024年第8期67卷 2592-2614页

作者： Qikai sun Fengge SU He sun State Key Laboratory of Tibetan Plateau Earth System Environment and Resources(TPESER)Institute of Tibetan Plateau ResearchChinese Academy of SciencesBeijing 100101China University of Chinese Academy of Sciences Beijing 100101China

The diverse climates,distribution of snow and glaciers,and geographic locations directly affect the runoff response to climate change in the upper basins of the Third *** present,a comprehensive analysis of runoff variations and their distinct responses to climate change in the westerlies-and monsoon-dominated upper basins is still *** study comprehensively analyzed annual runoff variations in westerlies-dominated basins(the upper basins of the Aksu(UAKS),Syr Darya(USRD),Yarkant(UYK),Hotan(UHT),Amu Darya(UAMD),and Indus(UI))and monsoon-dominated basins(the upper basins of the Yangtze(UYA),Yellow(UYE),Lancang(ULC),Nujiang(UNJ),and Yarlung Zangbo(UYZ))of the Third Pole from 1961 *** multi-source meteorological data and large-scale circulation factors,this study investigated the divergent responses of runoff in the upper basins to climate change,and explored the large-scale circulation mechanisms underlying runoff variations in these upper *** results showed that:(1)The annual runoff in the majority of upper basins(except for the UYE and UYZ)exhibited an increasing trend,and the annual runoff in the UAKS,UYK,and UI showed a significant increasing trend from1961 to *** annual runoff in the upper basins of the Third Pole changed abruptly from decreasing to increasing between the 1980s and 2000s,with the exception of the UYE.(2)The runoff in the monsoon-dominated upper basins has been controlled primarily by changes in precipitation over the past 55 *** contrast,the runoff in the westerlies-dominated upper basins exhibited three distinct long-term responses to climate change:temperature-dominated(UYK and UHT),precipitation-dominated(USRD and UAMD),and the combined influence of precipitation and temperature(UAKS and UI).Since the 1960s,the sensitivity of runoff to warm season temperature changes in the most westerlies-dominated upper basins has decreased,while the response of runoff to precipitation changes has intensified.(3)The stud

关键词： Upper basins of the Third Pole Climate change Westerlies and monsoon Runoff variations Large-scale circulation factors

同方期刊数据库

New perspectives on deep carbon cycling

在线全文

学校读者我要写书评

暂无评论

Science China Earth Sciences 2024年第8期67卷 2411-2422页

作者： Weidong sun Center of Deep Sea Research Key Laboratory of Ocean Observation and ForecastingInstitute of OceanologyChinese Academy of SciencesQingdao 266071China Laoshan Laboratory Qingdao 266237China University of Chinese Academy of Sciences Beijing 100049China

The proto-atmosphere serves as a crucial starting point for the carbon *** based on atmospheric data from Mars and Venus suggest that Earth's proto-atmosphere contained>110 bar of CO_(2)and>2.6 bar of *** protoatmosphere had over 1000 bar of water vapor during the magma ocean stage,assuming the proto-ocean had a volume of two oceans of *** this stage both water and carbon dioxide were in a supercritical state at the magma-atmosphere *** serpentinization reactions occurred due to rock-water interaction,producing abundant ***,nitrogen is reduced to ammonia,and carbon dioxide to methane,forming carbonate *** proto-atmosphere dominated by methane,ammonia,and hydrogen formed a significant amount of amino acids through *** process is essential not only to the origin of life,but also to the early carbon-nitrogen cycle on *** the Hadean eon,a large amount of CO_(2)was sequestered as carbonate and organic ***,it mainly entered the deep mantle through mantle overturn or *** the mantle transition zone,carbonate undergoes“Redox freezing”,where carbonate is reduced to diamond through oxidation of ferrous iron in the *** the lower mantle,Fe^(2+)undergoes disproportionation reactions,forming Fe^(3+)and metallic *** these,Fe^(3+)mainly resides in bridgmanite,thereby increasing the oxygen fugacity of the lower mantle,while metallic iron falls to the Earth's *** distribution of carbon in the mantle is crucial for deep carbon *** density curves of diamond and mantle peridotite melt intersect at the bottom of the mantle transition zone(about 660 km).This density crossover leads to the accumulation of diamond during the magma ocean *** materials such as subducting slabs enter the lower mantle,compensatory upwelling of lower mantle material *** enters the upper mantle,decomposes,releasing Fe^(3+)ions and oxidizes diamond

关键词： Carbon cycling Proto-atmosphere Amino acids Subduction initiation Passive margin subduction Atmospheric carbon dioxide

Fossil Equidae in the Linxia Basin with Biostratigraphic and Paleozoogeographic Significance

在线全文

学校读者我要写书评

暂无评论

Acta Geologica Sinica(English Edition) 2024年第1期98卷 1-9页

作者： sun Boyang Key Laboratory of Vertebrate Evolution and Human Origins of the Chinese Academy of Sciences Institute of Vertebrate Paleontology and PaleoanthropologyChinese Academy of SciencesBeijing 100044China

The Linxia Basin is characterized by an abundance of Cenozoic sediments,that contain exceptionally rich fossil *** are abundant in the Linxia Basin,the fossil record of equids in this region including 16 species that represent 10 *** comparison to other classic late Cenozoic areas in China,the Linxia Basin stands out,because the fauna and chronological data accompanying Linxia equids render them remarkably useful for *** anchitheriines in the region,such as Anchitherium and Sinohippus,represent early equids that appeared in the late stages of the middle and late Miocene,*** the equines,most species of Chinese hipparions have been identified in the Linxia Basin and some species of the genera Hipparion and Hippotherium have FAD records for ***,Equus eisenmannae is one of the earliest known species of Equus in the Old World and is well-represented at the Longdan *** species with precise geohistorical distributions can serve as standards for high-resolution chronological units within this *** at the eastern margin of the Tibetan Plateau and subject to considerable uplift,the Linxia Basin has served as a biogeographic transition area for equids throughout the late Cenozoic.

关键词： Equidae Cenozoic biostratigraphy paleozoogeography Linxia Basin