Please wait a minute...
工程设计学报  2024, Vol. 31 Issue (6): 689-696    DOI: 10.3785/j.issn.1006-754X.2024.03.407
【特约专栏】“双碳”背景下新型能源装备设计、制造、运维关键技术及其应用     
一种基于场景人物数量的任务卸载方案:针对云边协同智能监控系统
庞笛(),魏喆(),陈墨,张凯
沈阳工业大学 机械工程学院,辽宁 沈阳 110020
A task offloading scheme based on number of scene characters: for cloud edge collaborative intelligent monitoring system
Di PANG(),Zhe WEI(),Mo CHEN,Kai ZHANG
School of Mechanical Engineering, Shenyang University of Technology, Shenyang 110020, China
 全文: PDF(5162 KB)   HTML
摘要:

在云边协同智能监控系统的边缘计算设备上部署行为识别算法时,由于缺乏合理的任务卸载方案,使得系统的计算资源分配不均,从而导致系统运行功耗不稳定,并影响识别的速度及准确率。为解决上述问题,设计了一种基于场景人物数量的任务卸载方案,以优化云边协同智能监控系统的运行稳定性和识别效果。首先,对智能监控系统的运行参数进行了采集,确定了其功耗曲线和识别性能。然后,设计了轻量型人物数量识别模块,并通过程序编写实现了基于场景人物数量的监控任务分类。接着,测试了不同视频采样率对智能监控系统功耗和识别性能的影响,并确定了最佳采样率分配方案。最后,在复兴号动车组生产线的智能监控系统上对所提出的任务卸载方案进行了测试。结果显示,相较于现有的并行式任务卸载方案,基于场景人物数量的任务卸载方案使该生产线智能监控系统的平均识别准确率提高了0.53%,平均延迟缩短了1.56%,平均功耗降低了14.47%,有效地提高了系统的运行稳定性。研究结果对云边协同智能监控系统运行稳定性和识别效果的优化具有重要意义,可为其性能提升提供理论依据和工程支持。

关键词: 边缘计算智能监控系统任务卸载云边协同    
Abstract:

When the behavior recognition algorithm is deployed on the edge computing device of the cloud edge collaborative intelligent monitoring system, due to the lack of a reasonable task offloading scheme, the computing resources of the system are distributed unevenly, which leads to unstable system operation power consumption and affects the speed and accuracy of recognition. To solve the above problems, a task offloading scheme based on the number of scene characters has been designed to optimize the operational stability and recognition effect of the cloud edge collaborative intelligent monitoring system. Firstly, the operating parameters of the intelligent monitoring system were collected, and its power consumption curve and recognition performance were determined. Next, a lightweight character number recognition module was designed, and the classification of monitoring tasks based on the number of scene characters was realized by programming. Then, the influence of different video sampling rates on the power consumption and recognition performance of the intelligent monitoring system was tested, and the optimal sampling rate allocation scheme was determined. Finally, the proposed task offloading scheme was tested on the intelligent monitoring system for the production line of Fuxing electric multiple units. The results showed that compared with the existing parallel task offloading scheme, the task offloading scheme based on the number of scene characters improved the average recognition accuracy of the intelligent monitoring system of the production line by 0.53%, reduced average delay by 1.56%, and reduced average power consumption by 14.47%, which effectively improved the operational stability of the system. The research results are of great significance for optimizing the operational stability and recognition effect of the cloud edge collaborative intelligent monitoring system, and can provide theoretical basis and engineering support for its performance improvement.

Key words: edge computing    intelligent monitoring system    task offloading    cloud edge collaboration
收稿日期: 2023-12-06 出版日期: 2024-12-31
CLC:  X 924.3  
基金资助: 国家自然科学基金面上项目(51975386);辽宁省“揭榜挂帅”科技项目(2022020630-JH1/108);中国国家铁路集团有限公司科技研究开发计划资助项目(N2022J014)
通讯作者: 魏喆     E-mail: 13624766332@163.com;weizhe@sut.edu.cn
作者简介: 庞 笛(1997—),男,硕士生,从事智能制造、边缘计算研究,E-mail: 13624766332@163.com
服务  
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章  
庞笛
魏喆
陈墨
张凯

引用本文:

庞笛,魏喆,陈墨,张凯. 一种基于场景人物数量的任务卸载方案:针对云边协同智能监控系统[J]. 工程设计学报, 2024, 31(6): 689-696.

Di PANG,Zhe WEI,Mo CHEN,Kai ZHANG. A task offloading scheme based on number of scene characters: for cloud edge collaborative intelligent monitoring system[J]. Chinese Journal of Engineering Design, 2024, 31(6): 689-696.

链接本文:

https://www.zjujournals.com/gcsjxb/CN/10.3785/j.issn.1006-754X.2024.03.407        https://www.zjujournals.com/gcsjxb/CN/Y2024/V31/I6/689

图1  云边协同智能监控系统
图2  生产场景人员行为数据集
图3  智能监控系统的识别准确率
图4  智能监控系统的延迟
图5  智能监控系统的CPU、GPU功耗
图6  智能监控系统的总功耗
图7  场景人物数量识别模块
任务类型人数/人
停产状态0
正常状态1~5
特殊状态>5
表1  监控任务分类策略
图8  基于LRCN的行为识别算法流程
图9  不同采样率下智能监控系统的识别准确率
图10  不同采样率下智能监控系统的延迟
采样率/Hz

平均识别

准确率/%

平均延迟/

ms

平均功耗/

W

187.4599.43112.71
589.5099.51112.93
1092.2199.58113.59
1593.60101.80115.76
2094.21102.03116.57
2594.32106.17130.09
表2  不同采样率下智能监控系统的运行参数
图11  智能监控系统的识别准确率对比
图12  智能监控系统的延迟对比
图13  智能监控系统的功耗对比
智能监控系统平均识别准确率/%平均延迟/ms平均功耗/W功耗方差/W2
采用并行式任务卸载方案94.32106.17116.57145.32
采用本文任务卸载方案94.82104.5199.700.434 5
表3  智能监控系统的运行参数对比
1 阴艳超, 梁珉清, 张万达, 等. 基于云边协同的智能数控车间自调控系统研究与实现[J/OL]. 计算机集成制造系统: 1-18 (2023-04-10) [2023-10-22]. .
YIN Y C, LIANG M Q, ZHANG W D, et al. Research and implementation of intelligent NC workshop self-regulation system based on cloud-edge collaboration[J/OL]. Computer Integrated Manufacturing Systems: 1-18 (2023-04-10) [2023-10-22]. .
2 KIM S, PARK J, JEONG Y, et al. Intelligent monitoring system with privacy preservation based on edge AI[J]. Micromachines, 2023, 14(9): 1749.
3 HU L Y, SUN G D, REN Y L. CoEdge: exploiting the edge-cloud collaboration for faster deep learning[J]. IEEE Access, 2020, 8: 100533-100541.
4 WANG D Y, SUN X B, LIU Y, et al. Research on edge computing and caching resource allocation mechanism for multi-view video[C]//2022 10th International Conference on Intelligent Computing and Wireless Optical Communications. Chongqing, Jun. 10-12, 2022.
5 LI E, ZHOU Z, CHEN X. Edge intelligence: on-demand deep learning model co-inference with device-edge synergy[C]//Proceedings of the 2018 Workshop on Mobile Edge Communications. Budapest, Aug. 20, 2018.
6 NIKOUEI S Y, CHEN Y, SONG S J, et al. Real-time human detection as an edge service enabled by a lightweight CNN[C]//2018 IEEE International Conference on Edge Computing. San Francisco, CA, Jul. 2-7, 2018.
7 YAO S C, ZHAO Y R, ZHANG A, et al. DeepIoT: compressing deep neural network structures for sensing systems with a compressor-critic framework[C]// Proceedings of the 15th ACM Conference on Embedded Network Sensor Systems. Delft, Nov. 6-8, 2017.
8 李华, 赵领娣, 陈雨杰, 等. 多流融合的轻量级图卷积行为识别算法[J]. 计算机科学, 2023, 50(11A): 220800147-6.
LI H, ZHAO L D, CHEN Y J, et al. Lightweight graph convolution action recognition algorithm based on multi-stream fusion[J]. Computer Science, 2023, 50(11A): 220800147-6.
9 辛莉, 赵钦炎, 许茜, 等. 基于计算机视觉的复杂场景车牌识别算法[J]. 制造业自动化, 2021, 43(12): 135-139.
XIN L, ZHAO Q Y, XU Q, et al. License plate recognition algorithm in complex scene based on computer vision[J]. Manufacturing Automation, 2021, 43(12): 135-139.
10 LI Y J, ZHANG S K, WANG Z C, et al. TokenPose: learning keypoint tokens for human pose estimation[C]//2021 IEEE/CVF International Conference on Computer Vision. Montreal, QC, Oct. 10-17, 2021.
11 LI H, OTA K, DONG M X. Learning IoT in edge: deep learning for the Internet of Things with edge computing[J]. IEEE Network, 2018, 32(1): 96-101.
12 LIU C, CAO Y, LUO Y, et al. A new deep learning-based food recognition system for dietary assessment on an edge computing service infrastructure[J]. IEEE Transactions on Services Computing, 2018, 11(2): 249-261.
13 JIANG A H, WONG D L K, CANEL C, et al. Mainstream: dynamic stem-sharing for multi-tenant video processing[C]//Proceedings of the 2018 USENIX Annual Technical Conference. Boston, MA, Jul. 11-13, 2018.
14 TEERAPITTAYANON S, MCDANEL B, KUNG H T. Distributed deep neural networks over the cloud, the edge and end devices[C]//2017 IEEE 37th International Conference on Distributed Computing Systems. Atlanta, GA, Jun. 5-8, 2017.
15 MAO J C, CHEN X, NIXON K W, et al. MoDNN: local distributed mobile computing system for deep neural network[C]//Design, Automation & Test in Europe Conference & Exhibition. Lausanne, Mar. 27-31, 2017.
16 ZHAO Z R, BARIJOUGH K M, GERSTLAUER A. DeepThings: distributed adaptive deep learning inference on resource-constrained IoT edge clusters[J]. IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 2018, 37(11): 2348-2359.
17 DING C T, ZHOU A, LIU Y X, et al. A cloud-edge collaboration framework for cognitive service[J]. IEEE Transactions on Cloud Computing, 2022, 10(3): 1489-1499.
18 CHAKRABORTY S, MAZUMDAR K, DE D, et al. RMS: a delay sensitive road monitoring system using edge intelligence[J]. IEEE Sensors Journal, 2023, 23(3): 2643-2650.
19 HSU C W, HSU Y L, WEI H Y. Energy-efficient edge offloading in heterogeneous industrial IoT networks for factory of future[J]. IEEE Access, 2020, 8: 183035-183050.
20 BOUKERCHE A, GUAN S C, DE GRANDE R E. Sustainable offloading in mobile cloud computing[J]. ACM Computing Surveys, 2020, 52(1): 1-37.
21 WANG Z Y, CUI Y, LAI Z Q. A first look at mobile intelligence: architecture, experimentation and challenges[J]. IEEE Network, 2019, 33(4): 120-125.
22 CAI X Y, ZHOU W G, WU L, et al. Effective active skeleton representation for low latency human action recognition[J]. IEEE Transactions on Multimedia, 2016, 18(2): 141-154.
23 WU H Y, ALBIERO V, KRISHNAPRIYA K S, et al. Face recognition accuracy across demographics: shining a light into the problem[C]//2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops. Vancouver, BC, Jun. 17-24, 2023.
24 HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, Jun. 27-30, 2016.
No related articles found!