检索结果-南通市图书馆

Chinese Journal of Mechanical Engineering 2024年第4期37卷 215-230页

作者： Xiancheng Zhang Xin Pan Hao Zeng Haofu Zhou beijing Key Laboratoryof Health Monitoring and Self-reco.ery for High-end Mechanical Equipment Beijing University of Chemical TechnologyBeijing 100029China Engineering Research Center of Chemical Safety(Ministry of Education) Beijing University of Chemical TechnologyBeijing 100029China Engineering Management Department of China Three Gorges New Energy(Group)co. Ltd.Beijing 101199China

Intelligent fault diagnosis is an important method in rotating machinery fault diagnosis and equipment health *** deal with co.frequency vibration faults,a type of typical fault in rotating machinery,this paper proposes a fault diagnosis method based on the stacked autoenco.er(SAE)and ensembled ***,the time-and frequency-domain features of several co.frequency vibration faults are summarized based on the mechanism analysis and calculated using actual vibration *** realize and validate the high-precision diagnosis method of rotating equipment with co.frequency faults proposed in this study,the following three criteria are required:First,to improve the effectiveness and robustness of the ensembled model and the sliding window using data augmentation,adding noise,autoenco.er(AE)and SAE methods are analyzed in terms of principle and practical ***,ResNet is used as the feature extractor for the ensembled ResNet-SVM *** extraction is carried out twice,and the extracted co.frequency fault features are more ***,the data augmentation method and ensemble ResNet-SVM are co.bined for fault diagnosis and co.pared with other *** experimental results show that the accuracy of the proposed method can exceed 99.9%.

关键词： co.frequency vribation data argumentation Ensembeled ResNet-SVM High precision fault diagnosis

Text Difficulty Study:Do Machines Behave the Same as Humans Regarding Text Difficulty?

在线全文

学校读者我要写书评

暂无评论

Machine Intelligence Research 2024年第2期21卷 283-293页

作者： Bowen Chen Xiao Ding Yi Zhao Bo Fu Tingmao Lin Bing Qin Ting Liu Research Center for Social co.puting and Information Retrieval Harbin Institute of TechnologyHarbin 150001China Fundamental technology Center China Construction Bank Financial Technology Co.Ltd.Beijing 100033China

With the emergence of pre-trained models,current neural networks are able to give task performance that is co.parable to ***,we know little about the fundamental working mechanism of pre-trained models in which we do not know how they approach such performance and how the task is solved by the *** example,given a task,human learns from easy to hard,whereas the model learns ***,difficulty-insensitive learning leads to great success in natural language processing(NLP),but little attention has been paid to the effect of text difficulty in *** propose a human learning matching index(HLM Index)to investigate the effect of text *** results show:1)LSTM gives more human-like learning behavior than ***,UID-SuperLinear gives the best evaluation of text difficulty among four text difficulty *** nine tasks,some tasks’performance is related to text difficulty,whereas others are not.2)Model trained on easy data performs best in both easy and medium test data,whereas trained on hard data only performs well on hard test data.3)Train the model from easy to hard,leading to quicker co.vergence.

关键词： co.nition inspired natural language processing psycholinguistics explainability text difficulty curriculum learning

Hybrid model for BOF oxygen blowing time prediction based on oxygen balance mechanism and deep neural network

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Minerals,Metallurgy and Materials 2024年第1期31卷 106-117页

作者： Xin Shao Qing Liu Zicheng Xin Jiangshan Zhang Tao Zhou Shaoshuai Li State Key Laboratory of Advanced Metallurgy University of Science and Technology BeijingBeijing 100083China Engineering Research Center of MES technology for Iron&Steel Production Ministry of EducationBeijing 100083China Laiwu Iron and Steel Group Yinshan Section Steel co. Ltd.Jinan 271104China

The amount of oxygen blown into the co.verter is one of the key parameters for the co.trol of the co.verter blowing process,which directly affects the tap-to-tap time of co.verter. In this study, a hybrid model based on oxygen balance mechanism (OBM) and deep neural network (DNN) was established for predicting oxygen blowing time in co.verter. A three-step method was utilized in the hybrid model. First, the oxygen co.sumption volume was predicted by the OBM model and DNN model, respectively. Seco.d, a more accurate oxygen co.sumption volume was obtained by integrating the OBM model and DNN model. Finally, the co.verter oxygen blowing time was calculated acco.ding to the oxygen co.sumption volume and the oxygen supply intensity of each heat. The proposed hybrid model was verified using the actual data co.lected from an integrated steel plant in China, and co.pared with multiple linear regression model, OBM model, and neural network model including extreme learning machine, back propagation neural network, and DNN. The test results indicate that the hybrid model with a network structure of 3 hidden layer layers, 32-16-8 neurons per hidden layer, and 0.1 learning rate has the best prediction accuracy and stronger generalization ability co.pared with other models. The predicted hit ratio of oxygen co.sumption volume within the error±300 m^(3)is 96.67%;determination co.fficient (R^(2)) and root mean square error (RMSE) are0.6984 and 150.03 m^(3), respectively. The oxygen blow time prediction hit ratio within the error±0.6 min is 89.50%;R2and RMSE are0.9486 and 0.3592 min, respectively. As a result, the proposed model can effectively predict the oxygen co.sumption volume and oxygen blowing time in the co.verter.

关键词： basic oxygen furnace oxygen co.sumption oxygen blowing time oxygen balance mechanism deep neural network hybrid model

Expert Experience and data-Driven Based Hybrid Fault Diagnosis for High-SpeedWire Rod Finishing Mills

在线全文

学校读者我要写书评

暂无评论

co.puter Modeling in Engineering & Sciences 2024年第2期138卷 1827-1847页

作者： Cunsong Wang Ningze Tang Quanling Zhang Lixin Gao Haichen Yin Hao Peng Institute of Intelligent Manufacturing Nanjing Tech UniversityNanjing210009China beijing Gaohuahan Intelligent technology co. Ltd.Beijing100084China beijingWavelet Rhythm technology co. Ltd.Beijing100020China co.lege of Mechanical and Power Engineering Nanjing Tech UniversityNanjing211816China

The reliable operation of high-speed wire rod finishing mills is crucial in the steel production *** co.plex system-level equipment,it is difficult for high-speed wire rod finishing mills to realize fault location and real-time *** solve the above problems,an expert experience and data-driven-based hybrid fault diagnosis method for high-speed wire rod finishing mills is proposed in this ***,based on its mechanical structure,time and frequency domain analysis are improved in fault feature *** approach of co.bining virtual value,peak value with kurtosis value index,is adopted in time domain *** adjustment and side frequency analysis are proposed in frequency domain analysis to obtain accurate co.ponent characteristic frequency and its co.responding ***,acco.ding to time and frequency domain characteristics,fault location based on expert experience is proposed to get an accurate fault ***,the proposed method is implemented in the equipment intelligent diagnosis *** taking an equipment fault on site,for example,the effectiveness of the proposed method is illustrated in the system.

关键词： High-speed wire rod finishing mills expert experience data-driven fault diagnosis

Heterogeneous information phase space reco.struction and stability prediction of filling body–surrounding rock co.bination

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Minerals,Metallurgy and Materials 2024年第7期31卷 1500-1511页

作者： Dapeng Chen Shenghua Yin Weiguo Long Rongfu Yan Yufei Zhang Zepeng Yan Leiming Wang Wei Chen School of Civil and Resources Engineering University of Science and Technology of BeijingBeijing 100083China Production technology Department Longshou MineJinchuan Group Co.Ltd.Jinchang 737100China School of Civil Engineering and Architecture East China Jiaotong UniversityNanchang 330013China

Traditional research believes that the filling body can effectively co.trol stress co.centration while ignoring the problems of unknown stability and the co.plex and changeable stress distribution of the filling body–surrounding rock co.bination under high-stress *** monitoring data processing methods cannot fully co.sider the co.plexity of monitoring objects,the diversity of monitoring methods,and the dynamics of monitoring *** solve this problem,this paper proposes a phase space reco.struction and stability prediction method to process heterogeneous information of backfill–surrounding rock *** three-dimensional monitoring system of a large-area filling body–surrounding rock co.bination in Longshou Mine was co.structed by using drilling stress,multipoint displacement meter,and *** information,such as the stress and displacement of the filling body–surrounding rock co.bination,was co.tinuously *** with the average mutual information method and the false nearest neighbor point method,the phase space of the heterogeneous information of the filling body–surrounding rock co.bination was then *** this paper,the distance between the phase point and its nearest point was used as the index evaluation distance to evaluate the stability of the filling body–surrounding rock *** evaluated distances(ED)revealed a high sensitivity to the stability of the filling body–surrounding rock *** new method was then applied to calculate the time series of historically ED for 12 measuring points located at Longshou *** moments of mutation in these time series were at least 3 months ahead of the roadway return *** the ED prediction experiments,the autoregressive integrated moving average model showed a higher prediction accuracy than the deep learning models(long short-term memory and Transformer).Furthermore,the root-mean-square error distribution of the prediction results peaked at 0

关键词： deep mining filling body–surrounding rock co.bination phase space reco.struction multiple time series stability prediction

QoS Routing Optimization Based on Deep Reinforcement Learning in SDN

在线全文

学校读者我要写书评

暂无评论

co.puters, Materials & co.tinua 2024年第5期79卷 3007-3021页

作者： Yu Song Xusheng Qian Nan Zhang Wei Wang Ao Xiong State Key Laboratory of Networking and Switching technology Beijing University of Posts and TelecommunicationsBeijing100876China Marketing Service Center State Grid Jiangsu Electric Power Co.Ltd.Nanjing220000China Customer Service Center State Grid Co.Ltd.Nanjing211161China

To enhance the efficiency and expediency of issuing e-licenses within the power sector, we must co.front thechallenge of managing the surging demand for data traffic. Within this realm, the network imposes stringentQuality of Service (QoS) requirements, revealing the inadequacies of traditional routing allocation mechanismsin acco.modating such extensive data flows. In response to the imperative of handling a substantial influx of datarequests promptly and alleviating the co.straints of existing technologies and network co.gestion, we present anarchitecture forQoS routing optimizationwith in SoftwareDefinedNetwork (SDN), leveraging deep reinforcementlearning. This innovative approach entails the separation of SDN co.trol and transmission functionalities, centralizingco.trol over data forwardingwhile integrating deep reinforcement learning for informed routing decisions. Byfactoring in co.siderations such as delay, bandwidth, jitter rate, and packet loss rate, we design a reward function toguide theDeepDeterministic PolicyGradient (DDPG) algorithmin learning the optimal routing strategy to furnishsuperior QoS provision. In our empirical investigations, we juxtapose the performance of Deep ReinforcementLearning (DRL) against that of Shortest Path (SP) algorithms in terms of data packet transmission delay. Theexperimental simulation results show that our proposed algorithm has significant efficacy in reducing networkdelay and improving the overall transmission efficiency, which is superior to the traditional methods.

关键词： Deep reinforcement learning SDN route optimization QoS

Mechanical properties and energy evolutions of burst-prone co.l samples with holes and fillings

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of co.l Science & technology 2024年第3期11卷 176-189页

作者： Yukai Fu Yongzheng Wu Junchen Li Penghe Zhou Zhuoyue Sun Jie He CCTEG co.l Mining Research Institute Beijing100013China co.l Mining and Designing Department Tiandi Science and Technology Co.Ltd.Beijing100013China co.l Mining and Designing Branch China Coal Research InstituteBeijing100013China Resource Engineering Department China University of Mining and TechnologyBeijing100083China

During the mining process of impact-prone co.l seams,drilling pressure relief can reduce the impact propensity of the co.l seam,but it also reduces the integrity and strength of the co.l mass at the side of the ***,studying the mechanical properties and energy evolution rules of co.l samples co.taining holes and filled structures has certain practical significance for achieving co.rdinated co.trol of co.l mine rockburst disasters and the stability of roadway surrounding *** achieve this aim,seven types of burst-prone co.l samples were prepared and subject to uniaxial co.pression experiments with the aid of a TAW-3000 electro-hydraulic servo testing ***,the stress–strain curves,aco.stic emission signals,DIC strain fields and other data were co.lected during the ***,the failure modes and energy evolutions of samples with varying drilled hole sizes and filling materials were *** results show that the indexes related to burst propensity of the drilled co.l samples decline to some extent co.pared with those of the intact one,and the decline is positively co.elated to the diameter of the drilled *** hole filling,the strain co.centration degree around the drilled hole is lowered to a certain degree,and polyurethane filling has a more remarkable effect than cement ***,hole filling can enhance the strength and deformation resistance of *** drilling can accelerate the release of accumulated elastic strain energy,turning the aco.stic emission events from low-frequency and high-energy ones to high-frequency and low-energy ones,whereas hole filling can reduce the intensity of energy *** experimental results and theoretical derivation demonstrate that hole filling promotes co.l deformability and strength mainly by weakening stress co.centration surrounding the drilled ***,the fillings can achieve a better filling effect if their elastic modulus and Poisson’s ratio are closer

关键词： Rock mechanics co.l Mechanical properties Hole filling Energy evolution

Probing the electric double layer structure at nitrogen-doped graphite electrodes by co.stant-potential molecular dynamics simulations

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Energy Chemistry 2024年第6期93卷 299-305,I0008页

作者： Legeng Yu Nan Yao Yu-Chen Gao Zhong-Heng Fu Bo Jiang Ruiping Li Cheng Tang Xiang Chen Tsinghua Center for Green Chemical Engineering Electrification(CCEE) Department of Chemical EngineeringTsinghua UniversityBeijing 100084China beijing Advanced Innovation Center for Materials Genome Engineering Institute for Advanced Materials and TechnologyUniversity of Science and Technology BeijingBeijing 100083China National Materials co.rosion and Protection data Center University of Science and Technology BeijingBeijing 100083China Ordos Carbon Neutral Research and Application co. Ltd.Ordos 017010Inner Mongoli1China Institute for Carbon Neutrality Tsinghua UniversityBeijing 100084China Center for Next-Generation Energy Materials and School of Chemical Engineering Sungkyunkwan University2066 Seobu-roJangan-guSuwon-siGyeonggi-do 16419Republic of Korea

Electric double layer(EDL)is a critical topic in electrochemistry and largely determines the working performance of lithium ***,atomic insights into the EDL structures on heteroatom-modified graphite anodes and EDL evolution with electrode potential are very ***,a co.stant-potential molecular dynamics(CPMD)method is proposed to probe the EDL structure under working co.ditions,taking N-doped graphite electrodes and carbonate electrolytes as an *** interface model was developed,inco.porating the electrode potential and atom *** a result,an insightful atomic scenario for the EDL structure under varied electrode potentials has been established,which unveils the important role of doping sites in regulating both the EDL structures and the following electrochemical reactions at the atomic ***,the negatively charged N atoms repel the anions and adsorb Li~+at high and low potentials,*** preferential adsorption suggests that Ndoped graphite can promote Li~+desolvation and regulate the location of Li~+*** CPMD method not only unveils the mysterious function of N-doping from the viewpoint of EDL at the atomic level but also applies to probe the interfacial structure on other co.plicated electrodes.

关键词： Lithium batteries Graphite N-doping Electric double layer Molecular dynamics co.stant potential method Electrode potential

Disparity estimation for multi-scale multi-sensor fusion

在线全文

学校读者我要写书评

暂无评论

Journal of Systems Engineering and Electronics 2024年第2期35卷 259-274页

作者： SUN Guoliang PEI Shanshan LONG Qian ZHENG Sifa YANG Rui Suzhou Automotive Research Institute Tsinghua UniversitySuzhou 215000China beijing Smarter Eye technology co. Ltd.Beijing 100023China co.lege of Artifical Intelligence Tianjin University of Science and TechnologyTianjin 300457China State Key Laboratory of Automotive Safety and Energy Tsinghua UniversityBeijing 100084China Department of Precision Instrument Tsinghua UniversityBeijing 100084China

The perception module of advanced driver assistance systems plays a vital *** schemes often use a single sensor for data processing and environmental perception or adopt the information processing results of various sensors for the fusion of the detection *** paper proposes a multi-scale and multi-sensor data fusion strategy in the front end of perception and acco.plishes a multi-sensor function disparity map generation scheme.A binocular stereo vision sensor co.posed of two cameras and a light deterction and ranging(LiDAR)sensor is used to jointly perceive the environment,and a multi-scale fusion scheme is employed to improve the accuracy of the disparity *** solution not only has the advantages of dense perception of binocular stereo vision sensors but also co.siders the perception accuracy of LiDAR *** demonstrate that the multi-scale multi-sensor scheme proposed in this paper significantly improves disparity map estimation.

关键词： stereo vision light deterction and ranging(LiDAR) multi-sensor fusion multi-scale fusion disparity map

Method for the real-time detection of tomato ripeness using a phenotype robot and RP-YolactEdge

在线全文

学校读者我要写书评

暂无评论

International Journal of Agricultural and Biological Engineering 2024年第2期17卷 200-210页

作者： Yuanqiao Wang Wenbo Gou Chuanyu Wang Jiangchuan Fan Weiliang Wen Xianju Lu Xinyu Guo Chunjiang Zhao co.lege of Information Engineering Northwest A&F UniversityYangling 712100ShaanxiChina China National Engineering Research Center for Information technology in Agriculture(NERCITA) Beijing 100097China beijing Key Laboratory of Digital Plant Beijing Research Center for Information Technology in AgricultureBeijing 100097China beijing PAIDE Science and technology Development co. Ltd.Beijing 100097China

In order to address the challenge of non-destructive detection of tomato fruit ripeness in co.trolled environments,this study proposed a real-time instance segmentation method based on the edge *** method co.bined the principles of phenotype robots and machine vision based on deep learning.A co.pact and remotely co.trollable phenotype detection robot was employed to acquire precise data on tomato *** video data were then processed by using an efficient backbone and the FeatFlowNet structure for feature extraction and analysis of key-frame to non-key-frame mapping from video *** enhance the diversity of training datasets and the generalization of the model,an innovative approach was chosen by using random enhancement ***,the PolyLoss optimization technique was applied to further improve the accuracy of the ripeness multi-class detection *** validation,the method of this study achieved real-time processing speeds of 90.1 fps(RTX 3070Ti)and 65.5 fps(RTX 2060 S),with an average detection accuracy of 97%co.pared to manually measured *** is more accurate and efficient than other instance segmentation models acco.ding to actual testing in a ***,the results of this research can be deployed in edge devices and provide technical support for unmanned greenhouse monitoring devices or fruit-picking robots in facility environments.

关键词： instance segmentation phenotype robot tomato greenhouse-based plant phenotyping ripeness detection