|本期目录/Table of Contents|

[1]张晓华,马 煜,杨晨辉,等.基于卷积神经网络的设备安装位置智能识别方法[J].工业仪表与自动化装置,2021,(01):13-17.[doi:1000-0682(2021)01-0000-00]
　ZHANG Xiaohua,MA yu,YANG Chenhui,et al.Intelligent identification method of equipment installation position based on convolution neural network[J].Industrial Instrumentation & Automation,2021,(01):13-17.[doi:1000-0682(2021)01-0000-00]
点击复制

基于卷积神经网络的设备安装位置智能识别方法

《工业仪表与自动化装置》[ISSN:1000-0682/CN:61-1121/TH]

卷:
期数:
2021年01期
页码:
13-17
栏目:
出版日期:
2021-02-15

文章信息/Info

Title:
Intelligent identification method of equipment installation position based on convolution neural network
作者:
张晓华1; 马 煜2; 杨晨辉3; 隋雨杉3
1.国家电网东北电力调控分中心，辽宁 沈阳110180；
2.国网辽宁沈阳供电公司，辽宁 沈阳110000；
3.国网辽宁丹东供电公司，辽宁 丹东118000
Author(s):
ZHANG Xiaohua1; MA yu2; YANG Chenhui3; SUI Yushan3
1.State Grid Northeast Power Control Subcenter, Liaoning Shenyang 110180, China；
2.State Grid Shenyang Electric Power Supply Company, Liaoning Shenyang 110000,China；
3.State Grid Dandong Electric Power Supply Company, Liaoning Dandong 118000,China
关键词:
YOLO; 图像识别; 卷积神经网络; 深度学习
Keywords:
YOLO;  image recognition;  convolutional neural network;  deep learning
分类号:
TM715
DOI:
1000-0682(2021)01-0000-00
文献标志码:
A
摘要:
该文针对IT系统集成机柜建立了一套基于卷积神经网络的柜内设备智能识别系统。设计基于图像处理和深度学习的识别算法以实现机柜内设备安装位置的检测识别。建立设备图像数据集对YOLO模型进行训练，得到合适的加权系数，使模型损失函数值最小；将训练好的模型对实际环境中的机柜进行检测并与其他经典方法进行对比。实验结果显示，该系统检测的准确率和实时性都优于其他算法，准确率最高达到93.9%。对于建立基于机器视觉的设备识别和控制系统有重要意义，可以在一定程度是上取代人工识别和操作。
Abstract:
In this paper,a set of intelligent identification system based on convolution neural network is established for system integrated cabinet,and a recognition algorithm based on image processing and deep learning is designed to detect and identify the installation position of equipment in the cabinet.The equipment image data set is established to train the YOLO model,and the appropriate weighting coefficient is obtained to minimize the loss function value of the model.The trained model is used to detect the cabinet in the actual environment and compared with other classical methods. The experimental results show that the detection accuracy and real-time performance of the system are better than other algorithms.The highest accuracy is 93.9%.It is of great significance for the establishment of equipment identification and control system based on machine vision, which can replace manual identification and operation to a certain extent.

参考文献/References:

[1] SZEGEDY C, VANHOUCKE V, IOFFE S, et al. Rethinking the inception architecture for computer vision[C]. Proceedings of the IEEE conference on computer vision and pattern recognition, 2018:2818-2826.

[2] HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]. Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 770-778.
[3] 吴正文.卷积神经网络在图像分类中的应用研究[D].成都:电子科技大学, 2015.
[4] 吴进,钱雪忠.紧凑型深度卷积神经网络在图像识别中的应用[J].计算机科学与探索,2019,13(02):275-284.?/div>
[5] 杨珂,李从敏,周维勋,等.卷积神经网络多层特征联合的遥感图像检索[J].测绘科学,2019, 44(07):9-15+34.?/div>
[6] 邢世宏,施闻明,任荟洁.不同数据集容量下的卷积神经网络图像识别性能[J].舰船科学技术,2019,41(21): 188-193.
[7] 花如祥,吴国新,徐小力.卷积神经网络在图像识别中的优化研究[J].电子测量技术,2018,41(24):62-66.
[8] 林健巍.YOLO图像检测技术综述[J].福建电脑,2019(09): 80-83.
[9] 艾心.基于深度卷积神经网络的人脸识别研究[D].北京:北京邮电大学,2019.
[10] HE K M, ZHANG X, REN S Q, et al. Delving deep into rectifiers: surpassing human-level performance on image net classification[C]. IEEE International Conference on Computer Vision (ICCV),2015, 1026-1034.
[11] GLOROT X, BENGIO Y. Understanding the difficulty of training deep feedforward neural networks[C]. AISTATS, Sardinid, 2019:249-256.
[12] ZHUANG F Z, QI Z Y, DUAN K Y, et al. A comprehensive survey on transfer learning[J].Ar Xiv,2019, abs/1911. 02685.
[13] 汤高扬.基于MobileNet的目标跟踪算法研究与实现[D].广州:广东工业大学,2019.
[14] 杨炜,周凯霞,刘佳俊,等.结合迁移学习和Inception-V3模型的路面干湿状态识别方法[J].中国科技论文,2019, 14(08): 912-916.
[15] 张学典,顾璋琦,秦晓飞.基于VGG16模型的快速闭环检测算法[J].光学仪器,2019,41(03):20-26.
[16] SZEGEDY C, LIU W , JIA Y Q,et al. Going deeper with convolutions[C].IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, 2015:1-9.

相似文献/References:

[1]杨 帅,张有芬,李玉惠,等.基于深度卷积神经网络的车标分类[J].工业仪表与自动化装置,2017,(05):75.
　YANG Shuai,ZHANG Youfen,LI Yuhui,et al.Vehicle classification based on deep convolutional neural network[J].Industrial Instrumentation & Automation,2017,(01):75.
[2]郑昌庭,王 俊,郑 克.基于图像识别的变电站巡检机器人仪表识别研究[J].工业仪表与自动化装置,2020,(05):57.[doi:1000-0682(2020)05-0000-00]
　ZHENG Changting,WANG Jun,ZHENG Ke.Research on instrument recognition of substation inspection robot based on image recognition[J].Industrial Instrumentation & Automation,2020,(01):57.[doi:1000-0682(2020)05-0000-00]
[3]吕意飞,鲜 龙,文秀梅,等.变电站室内渗漏雨风险识别与实时告警装置的设计与应用[J].工业仪表与自动化装置,2020,(06):17.[doi:1000-0682(2020)06-0000-00]
　LYifei,XIAN Long,WEN Xiumei,et al.Design and application of risk identification and real-time warning device for indoor leakage rain in substation[J].Industrial Instrumentation & Automation,2020,(01):17.[doi:1000-0682(2020)06-0000-00]

备注/Memo

备注/Memo:
收稿日期：2020-07-30
作者简介：
张晓华（1963），男，辽宁昌图人，学士，高级工程师，主要研究方向为电力系统调度自动化。
更新日期/Last Update: 1900-01-01