检索结果-南通市图书馆

维普期刊数据库

A real-time Human Activity Recognition Approach with Generalization performance

在线全文

学校读者我要写书评

暂无评论

A Real-time Human Activity Recognition Approach with General...

Static Schedule Generation for time-Triggered Ethernet Based on Fuzzy Particle Swarm Optimization

第三十九届中国控制会议

作者： Shi-Jie Wei Bi Zhang Xiao-Wei Tan Xin-Gang Zhao Dan Ye College of Information Science and Engineering Northeastern University State Key Lab of Robotics Shenyang Institute of AutomationChinese Academy of Sciences Institutes of Robotics and Intelligent Manufacturing Chinese Academy of Sciences University of Chinese Academy of Sciences

The current human activity recognition（HAR） methods need training data from *** data collection causes discomfort to the users and most of the studies ignore the real-time performance of *** paper presents a real-time human activity recognition approach with strong generalization *** uses existing dataset to avoid long-term data collection of subjects,so that the machine can be quickly applied to each specific ***,it takes advantage of both combined accuracy and limited feature selection proposed by this paper to implement feature-selection-based transfer learning which improves HAR in both real-time and generalization *** view of the recognition time and accuracy,the depth neural network is selected,changeable structure of which is more suitable for feature *** approach utilizes four inertial measurement units placed on the outside of human thighs and shanks.A total of seven activities are taken into account that includes level-walking,upstairs,downstairs,uphill,downhill,standing and *** experiments are performed on six healthy male subjects in free-living settings to evaluate the efficacy of the *** approach achieved a notable activity recognition accuracy of 98.89%,and reported a fast average activity classification time of 28.6 ms.

关键词： Human Activity Recognition Feature Selection Generalization performance real-time performance Transfer Learning

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

Chinese Journal of Electronics 2019年第6期28卷 1250-1258页

作者： CHEN Huang WANG Lide SHEN Ping DI Jun School of Electrical Engineering Beijing Jiaotong University

In time-triggered ethernet(TTEthernet),designing and optimizing the static scheduling of TT messages to improve the real-time performance of the whole network control system is important. When there is a high load on TTEthernet network communication,the conventional static scheduling policy introduces certain problems, such as increased packet loss rate,load imbalance, and transmission delay. Meanwhile,basic artificial intelligence algorithms generally features convergence within only a few steps, leading to higher probability to fall into a local optimal solution. To realize these targets, Fuzzy particle swarm optimization(FPSO)based on population diversity is established. By combining Particle swarm optimization(PSO) with a fuzzy algorithm,setting the adjustment of the inertia weight and the regulation of the mutation factor as the controlled variables of the Fuzzy logic controller(FLC), and adjusting the FLC inference rules, a novel static schedule generation for TTEthernet is proposed. Simulation results prove that,compared to the conventional rate-monotonic scheduling algorithm and the PSO algorithm, FPSO has a strong global search capability when the communication of the TTEthernet system is in a high-load state. FPSO improves the load balancing of the network and reduces the transmission delay of TT message packets. FPSO shows excellent ability to optimize scheduling tables and guarantee the real-time performance of the TTEthernet system.

关键词： time-triggered ethernet(TTEthernet) Static schedule generation Fuzzy particle swarm optimization(FPSO) Fuzzy logic controller(FLC) Mutation factor real-time performance Load balancing Transmission delay

real-time and reliability analysis of time-triggered CAN-bus

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Aeronautics 2013年第1期26卷 171-178页

作者： Xia Jiqiang Zhang Chuansen Bai Ronggang Xue Liqiang School of Mechanical Engineering and Automation Beihang University

real-time performance and reliability are two most important issues in applications of time-triggered controller area network （CAN） bus systems at present. A scheduling matrix of time-triggered CAN-bus system is established using average-loading algorithm. Periodic messages are guaranteed to transmit without delay by distributing independent transmission windows within the system matrix. Considering the traditional CAN-bus transmission mechanism and the time-triggered feature, an algorithm is improved to calculate the worst-case delay of event-triggered messages in time-triggered CAN-bus systems. The failure probability is calculated for event-triggered messages whose worst-case delay exceeds their deadlines. Different levels of redundant structures of CAN-bus circuits are analyzed and the maintenance management is proposed to improve the system reliability. Finally, the reliabilities of different structures are calculated and the influences of maintenance on the system reliability are analyzed.

关键词： CAN-bus Maintenance system real-time performance Reliability time-triggered

维普期刊数据库

A Novel Gait Pattern Recognition Method Based on LSTM-CNN for Lower Limb Exoskeleton

在线全文

学校读者我要写书评

暂无评论

Journal of Bionic Engineering 2021年第5期18卷 1059-1072页

作者： Chao-feng Chen Zhi-jiang Du Long He Yong-jun Shi Jia-qi Wang Wei Dong State Key Laboratory of Robotics and System Harbin Institute of Technology(HIT)Harbin 150001China China South Industries Group Corporation Weapon Equipment Research InstituteBeijing 102202China

This paper describes a novel gait pattern recognition method based on Long Short-Term Memory(LSTM)and Convolutional Neural Network(CNN)for lower limb *** Inertial Measurement Unit(IMU)installed on the exoskeleton to collect motion information,which is used for LSTM-CNN *** article considers five common gait patterns,including walking,going up stairs,going down stairs,sitting down,and standing *** the LSTM-CNN model,the LSTM layer is used to process temporal sequences and the CNN layer is used to extract *** optimize the deep neural network structure proposed in this paper,some hyperparameter selection experiments were carried *** addition,to verify the superiority of the proposed recognition method,the method is compared with several common methods such as LSTM,CNN and *** results show that the average recognition accuracy can reach 97.78%,which has a good recognition eff ***,according to the experimental results of gait pattern switching,the proposed method can identify the switching gait pattern in time,which shows that it has good real-time performance.

关键词： Lower limb exoskeleton Gait pattern recognition LSTM-CNN Recognition accuracy real-time performance

Fast Iterative Closest Point-Simultaneous Localization and Mapping(ICP-SLAM)with Rough Alignment and Narrowing-Scale Nearby Searching

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 Donghua University(English Edition) 2017年第4期34卷 583-590页

作者：梁滨张金艺唐笛恺 Key Laboratory of Specialty Fiber Optics and Optical Access Networks School of Communication and Information EngineeringShanghai UniversityShanghai 200072China Microelectronic Research and Development Center Shanghai UniversityShanghai 200072China

Two deficiencies in traditional iterative closest pointsimultaneous localization and mapping( ICP-SLAM) usually result in poor real-time performance. On one hand, relative position between current scan frame and global map cannot be previously known. As a result, ICP algorithm will take much amount of iterations to reach convergence. On the other hand,establishment of correspondence is done by global searching, which requires enormous computational time. To overcome the two problems,a fast ICP-SLAM with rough alignment and narrowing-scale nearby searching is proposed. As for the decrease of iterative times,rough alignment based on initial pose matrix is proposed. In detail,initial pose matrix is obtained by micro-electro-mechanical system( MEMS) magnetometer and global landmarks. Then rough alignment will be applied between current scan frame and global map at the beginning of ICP algorithm with initial pose matrix. As for accelerating the establishment of correspondence, narrowingscale nearby searching with dynamic threshold is proposed,where match-points are found within a progressively constrictive *** to traditional ICP-SLAM,the experimental results show that the amount of iteration for ICP algorithm to reach convergence reduces to 92. 34% and ICP algorithm runtime reduces to 98. 86% on average. In addition,computational cost is kept in a stable level due to the eliminating of the accumulation of computational consumption. Moreover,great improvement can also been achieved in SLAM quality and robustness.

关键词： rough alignment initial pose matrix nearby searching dynamic threshold real-time performance

维普期刊数据库

Three-Dimensional Reconstruction of Complex Scenes of the Yangtze River Waterway

在线全文

学校读者我要写书评

暂无评论

Three-Dimensional Reconstruction of Complex Scenes of the Ya...

A Fast Target Tracking Algorithm Basted on Connected Component Labeling and Grey Value Statistics

作者： LI Zhi WANG Yi LI Caixia China Satellite Maritime Tracking and Control

The three-dimensional scene of the Yangtze River waterway is featured with perceptual intuition and easy understanding in spatial information compared with the traditional two-dimensional electronic waterway and will certainly play a key role in waterway planning,vessel monitoring,flood prevention and disaster *** analyzing the complexity of the Yangtze River waterway scenes,we propose the largescale three-dimensional reconstruction methods,illustrate the key technologies and solutions of improving the real-time performance and the three-dimensional visual effect of the system and prove the feasibility of the research method based on the detailed analysis of application examples.

关键词： the Yangtze River waterway 3D reconstruction real-time performance three-dimensional visual effect

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

A Fast Target Tracking Algorithm Basted on Connected Compone...

Implementation System of Human Eye Tracking Algorithm Based on FPGA

2012年计算机应用与系统建模国际会议

作者： Liu Qing Tang Linbo Zhao Baojun Sun Jingle School of Information and Electronics Beijing Inst. Of Technology

A fast target tracking algorithm based on connected component labeling and grey value statistics is proposed. First of all, nonlinear transformation is applied to perform enhancement for targets;and then adaptive threshold is applied to perform pre-processing division for the images;finally, target characteristics which are extracted through connected component labeling and grey value statistics are used to match the gray, the area and position of target, then pinpointing the target. Experimental results on the MATLAB software platform and the SOPC hardware platform show that it can perform real-time track accurately for the infrared small targets.

关键词： Infrared small target Connected component labeling Grey value statistics real-time performance

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

Computers, Materials & Continua 2019年第3期58卷 653-664页

作者： Zhong Liu Xin’an Wang Chengjun Sun Ken Lu Peking University Shenzhen Graduate SchoolsShenzhen518000China School of Electronics Engineering and Computer Science Peking UniversityeBeijing100871China iFutureLab Inc.955 Alma St. Suite BPalo AltoCA 94301USA.

With the high-speed development of transportation industry,highway traffic safety has become a considerable ***,with the development of embedded system and hardware chip,in recent years,human eye detection eye tracking and positioning technology have been more and more widely used in man-machine interaction,security access control and visual *** this paper,the high parallelism of FPGA was utilized to realize an elliptical approximate real-time human eye tracking system,which was achieved by the series register structure and random sample consensus(RANSAC),thus improving the speed of image processing without using external *** eye images acquired by the camera often generate a lot of noises due to uneven light and dark background,the preprocessing technologies such as color conversion,image filtering,histogram modification and image sharpening were *** terms of feature extraction of images,the eye tracking algorithm in this paper adopted seven-section rectangular eye tracking characteristic method,which increased a section between the mouth and the nose on the basis of the traditional six-section method,so its recognition accuracy is much *** is convenient for the realization of hardware parallel system in ***,aiming at the accuracy and real-time performance of the design system,a more comprehensive simulation test was carried *** human eye tracking system was verified on DE2-115 multimedia development platform,and the performance of VGA(resolution:640×480)images of 8-bit grayscale was *** results showed that the detection speed of this system was about 47 frames per second under the condition that the detection rate of human face(front face,no inclination)was 93%,which reached the real-time detection ***,the accuracy of eye tracking based on FPGA system was more than 95%,and it has achieved ideal results in real-time performance and robustness.

关键词： Human eye tracking FPGA real-time performance preprocessing elliptic approximation.