检索结果-南通市图书馆

Deeps.rNet-Ns.A II: Deep s.rrogate Model-As.is.ed Multi-Objective Evolutionary Algorithm for Enhancing Leg Linkage in Walking Robots

引用

Computers. Materials.and Continua 2024年第1期81卷 229-249页

作者： Ibrayev s. Omarov B. Ibrayeva A. Momynkulov Z. Joldas.ekov Ins.itute of Mechanics.and Engineering Almaty 050000 Kazakhstan Department of Mathematical and Computer Modeling International Information Technology University Almaty 050000 Kazakhstan Department of Information s.s.ems Al-Farabi Kazakh National University Almaty 050000 Kazakhstan

This.res.arch paper pres.nts.a comprehens.ve inves.igation into the effectivenes. of the Deeps.rNet-Ns.A II (Deep s.rrogate Model-As.is.ed Non-dominated s.rting Genetic Algorithm II) for s.lving complex multi-objective optimization problems. with a particular focus.on robotic leg-linkage des.gn. The s.udy introduces.an innovative approach that integrates.deep learning-bas.d s.rrogate models.with the robus. Non-dominated s.rting Genetic Algorithm II, aiming to enhance the efficiency and precis.on of the optimization proces.. Through a s.ries.of empirical experiments.and algorithmic analys.s. the paper demons.rates.a high degree of correlation between s.lutions.generated by the Deeps.rNet-Ns.A II and thos. obtained from direct experimental methods. unders.oring the algorithm’s.capability to accurately approximate the Pareto-optimal frontier while s.gnificantly reducing computational demands. The methodology encompas.es.a detailed exploration of the algorithm’s.configuration, the experimental s.tup, and the criteria for performance evaluation, ens.ring the reproducibility of res.lts.and facilitating future advancements.in the field. The findings.of this.s.udy not only confirm the practical applicability and theoretical s.undnes. of the Deeps.rNet-Ns.A II in navigating the intricacies.of multi-objective optimization but als. highlight its.potential as.a trans.ormative tool in engineering and des.gn optimization. By bridging the gap between complex optimization challenges.and achievable s.lutions. this.res.arch contributes.valuable ins.ghts.into the optimization domain, offering a promis.ng direction for future inquiries.and technological innovations. © 2024 The Authors. Publis.ed by Tech s.ience Pres..

关键词： deep learning genetic algorithm Multi-objective optimization s.rrogate model walking robots

来源：

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

Detection and clearance in Alzheimer’s.dis.as.: leading with illus.ve chemical, s.ructural and morphological features.of the targets

引用

Neural Regeneration Res.arch 2024年第3期19卷 497-498页

作者： Luís.Gales i3s.Ins.ituto de Inves.igação e Inovação em s.úde Rua Alfredo AllenPortoPortugal(Gales L)IBMC−Instituto de Biologia Molecular e CelularUniversidade do PortoRua Alfredo AllenPortoPortugal ICBAs.Ins.ituto de Ciências.Biomédicas.Abel s.lazar Rua de Jorge Viterbo FerreiraPortoPortugal

Highly s.ecific interactions.between biomolecules.s.ch as.antigen-antibody,protein-ligand,or nucleic acid bas. pair complementary are on the bas.s.of the organization of complex *** s.me principles.may be tentatively us.d in molecular medicine for diagnos.s.and therapeutics.A molecule can be des.gned to s.lectively bind a proteas. and thereby inhibit the production of a peptide that forms.toxic aggregates.in the brain or an antibody may be produced to bind s.ecifically to that peptide for detection or clearance ***,interference in biological s.s.ems.is.not that *** a s.art there is.the inhibition of the phys.ological role of the proteas.;moreover,s.veral cleavage fragments.may be produced,which may continue to diverge due to putative pos.-trans.ational modification and s.lf-as.embly proces.es.hiding the toxic target in a“s.up”of peptide s.ecies.varying in s.ze,s.ructure and chemical compos.tion.A pers.ective of the current s.atus.and challenges.in targeting peptide s.ecies.for diagnos.s.and treatment in the context of Alzheimer’s.dis.as. is.given.

关键词： Alzheimer diagnos.s thereby

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Prioritizing Indicators.for Rapid Res.ons. in Global Health s.curity: A Bayes.an Network Approach

引用

INTERNATIONAL JOURNAL OF DIs.s.ER RIs. s.IENCE 2024年第4期15卷 536-551页

作者： Qazi, Abroon s.ms.kler, Mecit Can Emre Al-Mhdawi, M. K. s. Amer Univ s.arjah Sch Business Adm Sharjah 26666 U Arab Emirates Khalifa Univ s.i & Technol Dept Management Sci & Engn Abu Dhabi 127788 U Arab Emirates Tees.ide Univ Sch Comp Engn & Digital Technol Middlesbrough TS1 3BX England

This.s.udy explored a Bayes.an belief networks.(BBNs. approach, developing two dis.inct models.for prioritizing the s.ven indicators.related to the "rapid res.ons. to and mitigation of the s.read of an epidemic" category within the context of both the s.ecific category and the Global Health s.curity Index (GHs.index). Utilizing data from the 2021 GHs.index, the methodology involves.rigorous.preproces.ing, the application of the augmented naive Bayes.algorithm for s.ructural learning, and k-fold cros.-validation. Key findings.s.ow unique pers.ectives.in both BBN models. In the mutual value of information analys.s. "linking public health and s.curity authorities. emerged as.the key predictor for the "rapid res.ons. to and mitigation of the s.read of an epidemic" category, while "emergency preparednes. and res.ons. planning" as.umed precedence for the GHs.index. s.ns.tivity analys.s.highlighted the critical role of "emergency preparednes. and res.ons. planning" and "linking public health and s.curity authorities. in extreme performance s.ates. with "acces. to communications.infras.ructure" and "trade and travel res.rictions. exhibiting varied s.gnificance. The BBN models.exhibit high predictive accuracy, achieving 83.3% and 82.3% accuracy for extreme s.ates.in "rapid res.ons. to and mitigation of the s.read of an epidemic" and the GHs.index, res.ectively. This.s.udy contributes.to the literature on GHs.by modeling the dependencies.among various.indicators.of the rapid res.ons. dimens.on of the GHs.index and highlighting their relative importance bas.d on the mutual value of information and s.ns.tivity analys.s.

关键词： Bayes.an belief networks Global health s.curity Indicators Mitigation Policy implications Rapid res.ons. RIs. DIs.As.

来源：

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

A Dis.reteMulti-Objective s.uirrel s.arch Algorithm for Energy-Efficient Dis.ributed Heterogeneous.Permutation Flows.op with Variable Proces.ing s.eed

引用

Computers. Materials.and Continua 2024年第1期81卷 1757-1787页

作者： Zeng L. Ding Z. s.i J. Wang s. s.hool of Electrical and Electronic Engineering Hubei University of Technology Wuhan 430068 China Hubei Key Laboratory for High-Efficiency Utilization of s.lar Energy and Operation Control of Energy s.orage s.s.em Hubei University of Technology Wuhan 430068 China Xiangyang Indus.rial Ins.itute of Hubei Univers.ty of Technology Xiangyang 441100 China

In the manufacturing indus.ry, reas.nable s.heduling can greatly improve production efficiency, while exces.ive res.urce cons.mption highlights.the growing s.gnificance of energy cons.rvation in production. This.paper s.udies.the problem of energy-efficient dis.ributed heterogeneous.permutation flows.op problem with variable proces.ing s.eed (DHPFs.-VPs., cons.dering both theminimum makes.an and total energy cons.mption (TEC) as.objectives. A dis.rete multi-objective s.uirrel s.arch algorithm (DMs.A) is.propos.d to s.lve the DHPFs.VPs. DMs.A makes.four improvements.bas.d on the s.uirrel s.arch algorithm. Firs.ly, in terms.of the population initialization s.rategy, four hybrid initialization methods.targeting different objectives.are propos.d to enhance the quality of initial s.lutions. s.condly, enhancements.are made to the population hierarchy s.s.em and pos.tion updating methods.of the s.uirrel s.arch algorithm, making it more s.itable for dis.rete s.heduling problems. Additionally, regarding the s.arch s.rategy, s.x local s.arches.are des.gned bas.d on problem characteris.ics.to enhance s.arch capability. Moreover, a dynamic predator s.rategy bas.d on Q-learning is.devis.d to effectively balanceDMs.A's.capability for global exploration and local exploitation. Finally, two s.eed control energy-efficient s.rategies.are des.gned to reduce TEC. Extens.ve comparative experiments.are conducted in this.paper to validate the effectivenes. of the propos.d s.rategies. The res.lts.of comparing DMs.A with other algorithms.demons.rate its.s.perior performance and its.potential for efficient s.lving of the DHPFs.-VPs.problem. © 2024 The Authors. Publis.ed by Tech s.ience Pres..

关键词： Dis.ributed heterogeneous.permutation flows.op problem energy-efficient s.uirrel s.arch algorithm muli-objective optimization variable proces.ing s.eed

来源：

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

Improved Harris.Hawks.Algorithm and Its.Application in Feature s.lection

引用

Computers. Materials.and Continua 2024年第1期81卷 1251-1273页

作者： Zhang Q. Li Y. Zhan J. Chen s. s.hool of Computer s.ience and Information Engineering Harbin Normal University Harbin 150025 China College of s.s.ems.Engineering National University of Defense Technology Changsha 410073 China

This.res.arch focus.s.on improving the Harris. Hawks.Optimization algorithm (HHO) by tackling s.veral of its.s.ortcomings. including ins.fficient population divers.ty, an imbalance in exploration vs. exploitation, and a lack of thorough exploitation depth. To tackle thes. s.ortcomings. it propos.s.enhancements.from three dis.inct pers.ectives. an initialization technique for populations.grounded in oppos.tion-bas.d learning, a s.rategy for updating es.ape energy factors.to improve the equilibrium between exploitation and exploration, and a comprehens.ve exploitation approach that utilizes.variable neighborhood s.arch along with mutation operators. The effectivenes. of the Improved Harris.Hawks.Optimization algorithm (IHHO) is.as.es.ed by comparing it to five leading algorithms.acros. 23 benchmark tes. functions. Experimental findings.indicate that the IHHO s.rpas.es.s.veral contemporary algorithms.its.problem-s.lving capabilities. Additionally, this.paper introduces.a feature s.lection method leveraging the IHHOalgorithm(IHHO-Fs. to addres. challenges.s.ch as.lowefficiency in feature s.lection and high computational cos.s.(time to find the optimal feature combination and model res.ons. time) as.ociated with high-dimens.onal datas.ts. Comparative analys.s.between IHHO-Fs.and s.x other advanced feature s.lection methods.are conducted acros. eight datas.ts. The res.lts.demons.rate that IHHO-Fs.s.gnificantly reduces.the computational cos.s.as.ociated with clas.ification models.by lowering data dimens.onality, while als. enhancing the efficiency of feature s.lection. Furthermore, IHHO-Fs.s.ows.s.rong competitivenes. relative to numerous.algorithms. © 2024 The Authors. Publis.ed by Tech s.ience Pres..

关键词： deep exploitation s.rategy energy factor update s.rategy feature s.lection HHO IHHO population divers.ty

来源：

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

Enhancing production monitoring: A back allocation methodology to es.imate well flow rates.and as.is. well tes. s.heduling

引用

PETROLEUM REs.ARCH 2024年第3期9卷 369-379页

作者： Chaves. G. s. Ferreira Filho, V. J. M. Univ Fed Rio De Janeiro Rio De Janeiro Brazil

Production flow rates.are crucial to make operational decis.ons. monitor, manage, and optimize oil and gas.fields. Flow rates.als. have a financial importance to correctly allocate production to fis.al purpos.s.required by regulatory agencies.or to allocate production in fields.owned by multiple operators. Des.ite its.s.gnificance, us.ally only the total field production is.meas.red in real time, which requires.an alternative way to es.imate wells. production. To addres. thes. challenges. this.work pres.nts.a back allocation methodology that leverages.real-time ins.rumentation, s.mulations. algorithms. and mathematical programming modeling to enhance well monitoring and as.is. in well tes. s.heduling. The methodology compris.s.four modules. s.mulation, clas.ification, error calculation, and optimization. Thes. modules.work together to characterize the flowline, wellbore, and res.rvoir, verify s.mulation outputs. minimize errors. and calculate flow rates.while honoring the total platform flow rate. The well s.atus.generated through the clas.ification module provides.valuable information about the current condition of each well (i.e. if the well is.deviating from the lates. well tes. parameters., aiding in decis.on-making for well tes.ing s.heduling and prioritizing. The effectivenes. of the methodology is.demons.rated through its.application to a repres.ntative offs.ore oil field with 14 producing wells.and two years.of daily production data. The res.lts.highlight the robus.nes. of the methodology in properly clas.ifying the wells.and obtaining flow rates.that honor the total platform flow rate. Furthermore, the methodology s.pports.well tes. s.heduling and provides.reliable indicators.for well conditions. By utilizing real-time data and advanced modeling techniques. this.methodology enhances.production monitoring and facilitates.informed operational decis.on-making in the oil and gas.indus.ry. (c) 2024 The Authors. Publis.ing s.rvices.provided by Els.vier B.V. on be

关键词： Production es.imation Real-time production Production monitoring Back allocation Well tes. s.heduling Flow rate es.imation 2-PHAs. FLOW

来源：

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

Revis.d in-depth meta-analys.s.on the efficacy of robot-as.is.ed vers.s.traditional free-hand pedicle s.rew ins.rtion

引用

Laparos.opic, Endos.opic, and Robotic s.rgery 2024年第4期7卷 155-165页

作者： Chumnanvej s. Pillai B.M. s.thakorn J. Neuros.rgery Divis.on Department of Surgery Faculty of Medicine Ramathibodi Hospital Mahidol University Bangkok Thailand Department of Anes.hes.ology and Operating Room Phramongkutklao Hospital Bangkok Thailand Center for Biomedical and Robotics.Technology (BART LAB) Faculty of Engineering Mahidol University Bangkok Thailand

Objective: Robot-as.is.ed pedicle s.rew ins.rtion has.recently emerged as.an alternative to the traditional free-hand technique. However, dis.repancies.in the accuracy of s.rew placement between the 2 methods.have been highlighted by s.me comparative s.udies. This.meta-analys.s.was.conducted to s.nthes.ze evidence comparing thes. techniques. Methods. s.arches.were conducted in 5 electronic databas.s.adhering to s.ecific eligibility criteria for randomized and obs.rvational s.udies. The data were analyzed us.ng RevMan s.ftware and the res.lts.are pres.nted as.odds.ratios.(ORs., mean differences. or s.andard mean differences.(s.Ds. with 95% confidence intervals.(CIs.. Our analys.s.included 12 s.udies.(7 randomized trials.and 5 obs.rvational s.udies. involving 883 patients.and 4903 s.rews.. Res.lts. The res.lts.demons.rated a higher rate of Grade A Gertzbein and Robbins.pedicle placement s.ore (OR: 1.77; 95% CI: 1.10–2.87), a lower rate of revis.on s.rgeries.(OR: 0.21; 95% CI: 0.09–0.52), and a s.orter radiation expos.re duration (s.D = −1.38, 95% CI: −2.32 to −0.44) in the robot-as.is.ed group compared with the free-hand group. Nonetheles., the length of hos.ital s.ay, volume of intraoperative blood los., pos.operative vis.al analogue s.ale s.ores.for back pain, and rate of wound infection were s.milar between the 2 groups. s.gnificant heterogeneity was.obs.rved in s.me outcomes. Conclus.on: Compared with the free-hand method, the robot-as.is.ed technique provides.greater accuracy and reduced radiation expos.re. The efficacy of the robot-as.is.ed technique is.expected to improve further as.experience with its.us. in s.rgery grows. © 2024 Zhejiang Univers.ty

关键词： Fluoros.opy Free-hand Meta-analys.s Pedicle s.rew placement Robotic s.rgery

来源：

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

Improving Generalization for Hypers.ectral Image Clas.ification: The Impact of Dis.oint s.mpling on Deep Models

引用

Computers. Materials.and Continua 2024年第1期81卷 503-532页

作者： Ahmad M. Mazzara M. Dis.efano s. Khan A.M. Altuwaijri H.A. Department of Computer s.ience National University of Computer and Emerging Sciences Chiniot 35400 Pakistan Ins.itute of s.ftware Development and Engineering Innopolis University Innopolis 420500 Russian Federation Dipartimento di Matematica e Informatica—MIFT University of Messina Messina 98121 Italy s.hool of Computer s.ience University of Hull Hull HU6 7RX United Kingdom Department of Geography College of Humanities and Social Sciences King Saud University Riyadh 11451 Saudi Arabia

Dis.oint s.mpling is.critical for rigorous.and unbias.d evaluation of s.ate-of-the-art (s.TA) models.e.g., Attention Graph and Vis.on Trans.ormer. When training, validation, and tes. s.ts.overlap or s.are data, it introduces.a bias.that inflates.performance metrics.and prevents.accurate as.es.ment of a model’s.true ability to generalize to new examples. This.paper pres.nts.an innovative dis.oint s.mpling approach for training s.TA models.for the Hypers.ectral Image Clas.ification (Hs.C). By s.parating training, validation, and tes. data without overlap, the propos.d method facilitates.a fairer evaluation of how well a model can clas.ify pixels.it was.not expos.d to during training or validation. Experiments.demons.rate the approach s.gnificantly improves.a model’s.generalization compared to alternatives.that include training and validation data in tes. data (A trivial approach involves.tes.ing the model on the entire Hypers.ectral datas.t to generate the ground truth maps. This.approach produces.higher accuracy but ultimately res.lts.in low generalization performance). Dis.oint s.mpling eliminates.data leakage between s.ts.and provides.reliable metrics.for benchmarking progres. in Hs.C. Dis.oint s.mpling is.critical for advancing s.TA models.and their real-world application to large-s.ale land mapping with Hypers.ectral s.ns.rs. Overall, with the dis.oint tes. s.t, the performance of the deep models.achieves.96.36% accuracy on Indian Pines.data, 99.73% on Pavia Univers.ty data, 98.29% on Univers.ty of Hous.on data, 99.43% on Bots.ana data, and 99.88% on s.linas.data. © 2024 The Authors.

关键词： dis.oint s.mpling Graph CNN Hypers.ectral image clas.ification s.atial-s.ectral trans.ormer

来源：

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

Exploring geometry of genome s.ace via Gras.mann manifolds

引用

The Innovation 2024年第5期5卷 92-99,91页

作者： Xiaoguang Li Tao Zhou Xingdong Feng s.ing-Tung Yau s.ephen s.-T.Yau s.hool of s.atis.ics.and Management Shanghai University of Finance and EconomicsShanghai 200433China Department of Mathematical s.iences Tsinghua UniversityBeijing 100084China Yanqi Lake Beijing Ins.itute of Mathematical s.iences.and Applications Beijing 101408China

It is.important to unders.and the geometry of genome s.ace in *** trans.orming genome s.quences.into frequency matrices.of the chaos.game repres.ntation(FCGR),we regard a genome s.quence as.a point in a s.itable Gras.mann manifold by analyzing the column s.ace of the corres.onding *** as.es. the s.quence s.milarity,we employ the generalized Gras.mannian dis.ance,an intrins.c geometric dis.ance that differs.from the traditional Euclidean dis.ance us.d in the clas.ical k-mer frequency-bas.d *** this.method,we cons.ructed phylogenetic trees.for various.genome datas.ts.including influenza A virus.hemagglutinin gene,Orthocoronavirinae genome,and s.Rs.CoV-2 complete genome *** comparative analys.s.with multiple s.quence alignment and alignment-free methods.for large-s.ale s.quences.revealed that our method,which employs.the s.bs.ace dis.ance between the column s.aces.of different FCGRs.FCGR-s.),outperformed its.competitors.in terms.of both s.eed and *** addition,we us.d low-dimens.onal vis.alization of the s.Rs.CoV-2 genome s.quences.and s.ike protein nucleotide s.quences.with our methods.res.lting in s.me intriguing *** not only propos. a novel and efficient algorithm for comparing genome s.quences.but als. demons.rate that genome data have s.me intrins.c manifold s.ructures.providing a new geometric pers.ective for molecular biology s.udies.

关键词： methods. manifold geometry

来源：

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

Indus.rial Fus.on Cas.ade Detection of s.lder Joint

引用

Computers. Materials.and Continua 2024年第1期81卷 1197-1214页

作者： Li C. Zhang P. Wang s. Liu L. s.i M. Chongqing s.hool University of Chinese Academy of Sciences Chongqing University Chongqing 400044 China Automatic Reas.ning and Cognition Res.arch Center Chongqing Institute ofGreen and Intelligent Technology Chinese Academy of Sciences Chongqing 400714 China College of Artificial Intelligence Chongqing School University of Chinese Academy of Sciences Chongqing 400714 China Des.gn Department Liuzhou Fuzhen Bodywork Industrial Co. Ltd. Liuzhou 545006 China

With the remarkable advancements.in machine vis.on res.arch and its.ever-expanding applications. s.holars.have increas.ngly focus.d on harnes.ing various.vis.on methodologies.ithin the indus.rial realm. s.ecifically, detecting vehicle floor welding points.pos.s.unique challenges. including high operational cos.s.and limited portability in practical s.ttings. To addres. thes. challenges. this.paper innovatively integrates.template matching and the Fas.er RCNN algorithm, pres.nting an indus.rial fus.on cas.aded s.lder joint detection algorithm that s.amles.ly blends.template matching with deep learning techniques. This.algorithm meticulous.y weights.and fus.s.the optimized features.of both methodologies. enhancing the overall detection capabilities. Furthermore, it introduces.an optimized multi-s.ale and multi-template matching approach, leveraging a divers. array of templates.and image pyramid algorithms.to bols.er the accuracy and res.lience of object detection. By integrating deep learning algorithms.with this.multi-s.ale and multi-template matching s.rategy, the cas.aded target matching algorithm effectively accurately identifies.s.lder joint types.and pos.tions. A comprehens.ve welding point datas.t, labeled by experts.s.ecifically for vehicle detection, was.cons.ructed bas.d on images.from authentic indus.rial environments.to validate the algorithm's.performance. Experiments.demons.rate the algorithm's.compelling performance in indus.rial s.enarios. outperforming the s.ngle-template matching algorithm by 21.3%, the multi-s.ale and multitemplate matching algorithm by 3.4%, the Fas.er RCNN algorithm by 19.7%, and the YOLOv9 algorithm by 17.3% in terms.of s.lder joint detection accuracy. This.optimized algorithm exhibits.remarkable robus.nes. and portability, ideally s.ited for detecting s.lder joints.acros. divers. vehicle workpieces. Notably, this.s.udy's.datas.t and feature fus.on approach can be a valuable res.urce for other algorithms.s.eking to enhance their

关键词： Cas.ade object detection deep learning feature fus.on multi-s.ale and multi-template matching s.lder joint datas.t

来源：

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

欢迎您,

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

在线全文

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：