检索结果-南通市图书馆

An improved defect reco.nition framework for casting based on DETR algorithm

在线全文

学校读者我要写书评

暂无评论

Journal of Iron and Steel Research International 2023年第5期30卷 949-959页

作者： Long Zhang Sai-fei Yan Jun Hong Qian Xie Fei Zhou Song-lin Ran anhui Province Key Laboratory of Metallurgical Engineering&Resources Recycling Anhui University of TechnologyMa’anshan 243002AnhuiChina School of Metallurgical Engineering Anhui University of TechnologyMa’anshan 243002AnhuiChina Technical Department Anhui Highly Precision Casting Co.Ltd.Ma’anshan 238100AnhuiChina anhui Engineering Laboratory for Intelligent Applications and Security of Industrial Internet Anhui University of TechnologyMa’anshan 243032AnhuiChina

The current casting surface defect detection algorithms suffer from poor small target defect reco.nition and imbalance between detection performance and detection *** improved algorithmic framework for casting defect detection was proposed based on the DEtection TRansformer(DETR)*** algorithm takes ResNet with an efficient channel attention(ECA)-Net module as the backbone *** addition,based on the original algorithm architecture,dynamic anchor boxes,improved multi-scale deformable attention module,and SIoU loss function are introduced to improve the sensitivity of transformer structure to input location information and scale size,and the small target defect detection performance is effectively *** reco.nition performance of the algorithm in a self-built casting defect dataset was *** improved DETR algorithm has 97.561% accuracy in reco.nizing two defects,namely sandinclusion and notch,with the detection rate being improved by 65.854% and 17.073% co.pared with the original DETR and you only look once(Yolo)-V5,*** algorithm verifies the applicability of the transformer architecture target detection algorithm for casting defect detection tasks and provides new ideas for detecting other similar application scenarios.

关键词： Casting defect reco.nition DEtection TRansformer Small target detection Deep learning Attention mechanism

Intelligent power distribution live‐line operation robot systems based on stereo camera

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

High Voltage 2023年第6期8卷 1306-1318页

作者： Yutao Chen Yahao Wang Xuming Tang Kai Wu Shaolei Wu Rui Guo Yu Feng Erbao Dong Department of precision Machinery and precision Instrumentation University of Science and Technology of ChinaHefeiChina State Grid anhui Electric Power co.pany HefeiChina State Grid anhui Electric Power co.pany Electric Power Research Institute HefeiChina State Grid Intelligent technology co. LtdJinanChina

Maintenance tasks in distribution networks are often acco.panied by hazards associated with high altitudes and high *** utilising robots instead of human operators to perform these tasks,potential risks can be avoided,while productivity can be *** research proposes an intelligent power distribution live‐line operation robot(PDLOR)system based on a stereo camera to replace human to co.plete *** PDLOR system co.sists of several key co.ponents,including dual manipulators,wireless tools,a visual perception system,an insulated bucket truck,and a ground co.trol *** the task is co.firmed,the real‐time vision system identification and posi-tioning enable the adjustment of the insulated bucket to position the robot co.rectly for its intended *** stereo camera plays a crucial role in accurately reco.nising and estimating the object's ***,a simplified reco.struction is performed within a virtual simulation environment,which aids in co.lision detection during path *** obtaining the optimal path,it is then co.municated to the real manip-ulator for *** validate the feasibility of the PDLOR system,field experiments were co.ducted in actual distribution network *** results demonstrate that the PDLOR effectively co.pletes single‐phase power‐line co.nection tasks within a remarkable 10‐min timeframe.

关键词： robot distribution stereo

Progress and prospects of pH-neutral aqueous organic redox flow batteries:Electrolytes and membranes

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

Journal of Energy Chemistry 2024年第9期96卷 89-109页

作者： Kang Peng Gonggen Tang Chao Zhang Xian Yang Peipei Zuo Zhanfeng Xiang Zhong Yao Zhengjin Yang Tongwen Xu Key Laboratory of precision and Intelligent Chemistry Department of Applied ChemistrySchool of Chemistry and Materials ScienceUniversity of Science and Technology of ChinaHefei 230026AnhuiChina Suqian Time Energy Storage technology co. Ltd.Suqian 223800JiangsuChina

Aqueous organic redox flow batteries(AORFBs),which exploit the reversible electrochemical reactions of water-soluble organic electrolytes to store electricity,have emerged as an efficient electrochemical energy storage technology for the grid-scale integration of renewable ***-neutral AORFBs that feature high safety,low co.rosivity,and environmental benignity are particularly promising,and their battery performance is significantly impacted by redox-active molecules and ion-exchange membranes(IEMs).Here,representative anolytes and catholytes engineered for use in pH-neutral AORFBs are outlined and summarized,as well as their side reactions that cause irreversible battery capacity *** addition,the recent achievements of IEMs for pH-neutral AORFBs are discussed,with a focus on the co.struction and tuning of ion transport ***,the critical challenges and potential research opportunities for developing practically relevant pH-neutral AORFBs are presented.

关键词： Aqueous organic redox flow battery pH-Neutral Anolyte Catholyte Membrane

Quantum key distribution system against the probabilistic faint after-gate attack

在线全文

学校读者我要写书评

暂无评论

co.munications in Theoretical Physics 2020年第11期72卷 43-47页

作者： Meng Ye Jian-Hui Li Yong Wang Peng Gao Xin-Xin Lu Yong-Jun Qian Power Generation co.pany GuangZhou 510700China anhui Qasky Quantum technology co. Ltd.Wuhu 241102China

In practical quantum key distribution(QKD)systems,a single photon-detector(SPD)is one of the most vulnerable *** after-gate attack is a universal attack against the ***,the original faint after-gate attack can be disco.ered by monitoring the *** paper presents a probabilistic generalization of the attack,which we refer to as probabilistic faint after-gate attack,by introducing probability co.trol *** co.ntermeasures for photocurrent monitoring may fail in detecting the eavesdropper under some specific *** mitigate this threat,we provide a method to determine the detectable boundary in the limitation of precision of photocurrent monitoring,and investigate the security of QKD systems under such boundaries using the weak randomness model.

关键词： quantum key distribution practical security single photon-detector weak randomness model

DLA+: A Light Aggregation Network for Object Classification and Detection

在线全文

学校读者我要写书评

暂无评论

International Journal of Automation and co.puting 2021年第6期18卷 963-972页

作者： Fu-Tian Wang Li Yang Jin Tang Si-Bao Chen Xin Wang School of co.puter Science and technology Anhui UniversityHefei 230601China Shenzhen Raixun Information technology co. Ltd.Shenzhen 518000China Peking University Shenzhen Institute Shenzhen 518000China

An efficient co.volution neural network(CNN) plays a crucial role in various visual tasks like object classification or detection, etc. The most co.mon way to co.struct a CNN is stacking the same co.volution block or co.plex co.nection. These approaches may be efficient but the parameter size and co.putation(co.p) have explosive growth. So we present a novel architecture called"DLA+", which co.ld obtain the feature from the different stages, and by the newly designed co.volution block, co.ld achieve better accuracy, while also dropping the co.putation six times co.pared to the baseline. We design some experiments about classification and object detection. On the CIFAR10 and VOC data-sets, we get better precision and faster speed than other architecture. The lightweight network even allows us to deploy to some low-performance device like drone, laptop, etc.

关键词： Light weight image classification channel attention efficient co.volution object detection

Judgment of blast furnace iron-tapping status based on data differential processing and dynamic window analysis algorithm

在线全文

学校读者我要写书评

暂无评论

Progress in Natural Science:Materials International 2023年第4期33卷 450-457页

作者： Youcheng Zong Zixian Wang Xiang Liu Yi Nian Jianian Pan Chaojie Zhang Yingxue Wang Jianhua Chu Liqiang Zhang The Intelligent Metallurgy Research Institute School of Metallurgical Engineering Anhui University of Technology Department of Technical Quality Manufacturing Management Division(Technical Center) Anhui Changjiang Steel Co. Ltd. National Engineering Research Center for Public Safety Risk Perception and co.trol by Big Data(RPP) China Academic of Electronics and Information Technology

Ironmaking plants hold a co.e position in steel production. Efficient level detection and iron-tapping status judgment can accurately measure the liquid level and predict the amount of molten iron in the ladle, thereby ensuring production safety and improving production efficiency. At present, the judgment of the start and end statuses of blast furnace iron-tapping mainly relies on manual visual inspection. However, due to the high-temperature environment on-site and interference from iron powder, gunning mud, and other dust, operators find it challenging to make accurate judgments. To address these issues, this study proposes a robust and accurate method for judging the iron-tapping status of blast furnaces. This study co.lected four million on-site liquid level data, underwent interpolation, noise reduction, normalization, and other preprocessing steps, and then used differential processing to calculate the iron-tapping rate. co.bined with dynamic window feature analysis, it effectively captured the change characteristics of the liquid level in the ladle under different iron-tapping *** simulated on-site experiments, the method achieved a 96.6 % accuracy rate for iron-tapping status *** real-time iron-tapping status judgment method based on differential processing of liquid level data and dynamic window feature analysis proposed in this study demonstrates high precision and robustness in the co.plex environment of ironmaking plants. This technology not only improves the production efficiency of the iron and steel link but also frees on-site operators from the harsh working environment, eliminating human errors in judgments and omissions, further promoting the intelligent process of ironmaking plants.

关键词： Blast furnace iron-tapping Molten iron level Radar level gauge Iron-tapping status judgment Data differential processing Sliding window

Deep Reinforcement Learning Based Resource Allocation for Fault Detection with Cloud Edge co.laboration in Smart Grid

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

CSEE Journal of Power and Energy Systems 2024年第3期10卷 1220-1230页

作者： Qiyue Li Yadong Zhu Jinjin Ding Weitao Li Wei Sun Lijian Ding Hefei University of technology Engineering technology Research Center of Industrial Automation Hefei 230009China Electric Power Research Institute of State Grid anhui Electric Power co. Ltd.Hefei 230000China Hefei University of technology Hefei 230009China IEEE CSEE

Real-time fault detection is important for operation of smart *** has beco.e a trend of future development to design an anomaly detection system based on deep learning by using the powerful co.puting power of the ***,delay of Internet transmission is large,which may make the delay time of detection and transmission go beyond the ***,the edge-based scheme may not be able to undertake all data detection tasks due to limited co.puting resources of edge ***,we propose a cloud-edge co.laborative smart grid fault detection system,next to which edge devices are placed,and equipped with a lightweight neural network with different precision for fault *** addition,a sub-optimal and realtime co.munication and co.puting resource allocation method is proposed based on deep reinforcement *** method greatly speeds up solution time,which can meet the requirements of data transmission delay,maximize the system throughput,and improve co.munication *** results show the scheme is superior in transmission delay and improves real-time performance of the smart grid detection system.

关键词： Cloud-edge co.laboration co.munication delay deep reinforcement learning fault detection smart grid

Data-driven casting defect prediction model for sand casting based on random forest classification algorithm

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

China Foundry 2024年第2期21卷 137-146页

作者： Bang Guan Dong-hong Wang Da Shu Shou-qin Zhu Xiao-yuan Ji Bao-de Sun Shanghai Key Lab of Advanced High-Temperature Materials and precision Forming School of Materials Science and EngineeringShanghai Jiao Tong UniversityShanghai 200240China State Key Lab of Metal Matrix co.posites Shanghai Jiao Tong UniversityShanghai 200240China Hefei Casting and Forging Factory of anhui Heli co. LtdHefei 230000China State Key Laboratory of Materials Processing and Die&Mould technology Huazhong University of Science and TechnologyWuhan 430074China

The co.plex sand-casting process co.bined with the interactions between process parameters makes it difficult to co.trol the casting quality,resulting in a high scrap rate.A strategy based on a data-driven model was proposed to reduce casting defects and improve production efficiency,which includes the random forest(RF)classification model,the feature importance analysis,and the process parameters optimization with Monte Carlo *** co.lected data includes four types of defects and co.responding process parameters were used to co.struct the RF *** results show a recall rate above 90% for all *** Gini Index was used to assess the importance of the process parameters in the formation of various defects in the RF ***,the classification model was applied to different production co.ditions for quality *** the case of process parameters optimization for gas porosity defects,this model serves as an experimental process in the Monte Carlo method to estimate a better temperature *** prediction model,when applied to the factory,greatly improved the efficiency of defect *** show that the scrap rate decreased from 10.16% to 6.68%.

关键词： sand casting process data-driven method classification model quality prediction feature importance

Process-based deep learning model:3D prediction method for shot peen forming of an aircraft panel

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2023年第11期36卷 500-514页

作者： Ziyu WANG Peng ZHANG Qun ZHANG Lijuan ZHOU Raneen Abd ALI Wenliang CHEN Lingling XIE co.lege of Mechanical and Electrical Engineering Nanjing University of Aeronautics and AstronauticsNanjing 210016China School of Mechanical Engineering Anhui University of TechnologyMa’anshan 243032China AVIC XAC Aerostructure(Hanzhong)Manufacturing co. LtdHanzhong 723000China Air co.ditioning and Refrigeration Techniques Engineering Department Al-Mustaqbal University CollegeBabylon 51001Iraq School of Metallurgical Engineering Anhui University of TechnologyMa’anshan 243032China

Shot peen-forming is a more precise method of forming aircraft panels than co.ventional *** traditional method of acquiring the process parameters relies mainly on prior theoretical knowledge and *** the finite element method’s ability to replace some experimentation,it still cannot realize the design of shot peen forming processes parameters of an aircraft panel based on a known *** study uses an innovative model-based deep learning approach to predict aircraft panel deformation and active design the shot peening *** prediction time is less than 1 seco.d,resulting in a significant reduction in co.putational *** shot peen forming process parameters and the geometric structure characteristics of the aircraft panel are divided into independent channels to establish a high-dimensional feature map,which are used to train the deep learning *** forming co.tours of the 2024-T351 high-strength aluminum alloy panel are predicted under different shot peening *** addition,the process parameters are designed acco.ding to the known co.tour of the forming *** verify the precision of the proposed method,the designed shot peen forming process is used to manufacture a single curvature aircraft panel with a curvature radius of 3500 *** is good agreement between the forming co.tour and the theoretical design *** maximum deformation error is less than 1 mm and its mean error is 7.8%.The mean curvature radius error is 5.668%.The proposed method provides a new and practical reference to the precise design of the shot peen-forming process.

关键词： Aircraft panel Deep learning Finite element method Process reverse design Shot peen forming