|本期目录/Table of Contents|

[1]苏 杨,余 萱.基于卷积神经网络的机房柜门开关状态识别研究[J].工业仪表与自动化装置,2019,(05 总269):22-25.[doi:1000-0682(2019)05-0000-00]
 SU Yang,YU Xuan.Study on state recognition of cabinet door switch based on convolutional neural network[J].Industrial Instrumentation & Automation,2019,(05 总269):22-25.[doi:1000-0682(2019)05-0000-00]
点击复制

基于卷积神经网络的机房柜门开关状态识别研究

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

卷:
期数:
2019年05期 总第269期
页码:
22-25
栏目:
出版日期:
2019-10-15

文章信息/Info

Title:
Study on state recognition of cabinet door switch based on convolutional neural network
作者:
苏 杨余 萱
贵州电网有限责任公司 信息中心,贵阳 550002
Author(s):
SU YangYU Xuan
( Information Center of Guizhou Power Grid Co., Ltd., Guiyang 550002, China)
关键词:
关键词:卷积神经网络数据扩充迁移学习批标准化
Keywords:
Keywords:convolutional neural networkdata expansion migration learning batch normalization
分类号:
TP391
DOI:
1000-0682(2019)05-0000-00
文献标志码:
A
摘要:
为监测电网系统中机房柜门开关状态,避免产生安全隐患和财产损失,提出了一种基于卷积神经网络的机房柜门开关状态识别方法。该方法以ZF卷积模型作为基础网络,通过引入数据扩充防止过拟合现象产生,并使用迁移学习及BN批标准化方法加快网络收敛速度,提升了识别精度。以电网系统机房中的机柜门开关状态图像作为分类对象进行实验。实验结果表明,该方法能避免人工提取特征的局限性,识别精度满足实际监控要求。
Abstract:
To monitor the switch status of cabinet doors in the power? system, avoiding potential safety hazards and property losses, a method based on convolutional neural network is presented. This method takes the convolution model of ZFNet as the basic network, preventing overfitting by introducing data expansion, speeding up network convergence and improving the recognition accuracy by using migration learning and batch normalization. The typical image of the status of the cabinet in the power grid system are used as the classification objects. Experimental results show that this method can avoid the limitation of manual feature extraction and the recognition accuracy meets the actual monitoring requirements.

参考文献/References:

[1] Li B H,Hou B C,Yu W T,et al.Applications of artificial intelligence in intelligent manufacturing: a review[J]. Frontiers of Information Technology & Electronic Engineering, 2017,18(01):86-96.

[2] 周昊飞,刘玉敏.基于小波重构的动态过程LSSVM- BPNN在线智能监控模型[J].系统管理学报,2018, 27(02): 291-298.
[3] 阚宝朋.基于PCA-SIFT的煤矿监控目标识别及行为分析[J].煤炭技术,2017,36(12):230-232.
[4] 苑洋,黄迪,王蕴红.面向不同距离的实时人体检测与跟踪系统[J].模式识别与人工智能,2014,27(10): 939-945.
[5] 卢宏涛,张秦川.深度卷积神经网络在计算机视觉中的应用研究综述[J].数据采集与处理,2016,31(01): 1-17.
[6] 李彦冬,郝宗波,雷航.卷积神经网络研究综述[J].计算机应用,2016,36(09): 2508-2515.
[7] Cui X, Goel V, Kingbury B. Data Augmentation for Deep Neural Network Acoustic Modeling[J]. IEEE/ ACM Transactions on Audio Speech & Language Processing, 2015, 23(09):1469-1477.
[8] Ioffe S, Szegedy C. Batch normalization: accelerating deep network training by reducing internal covariate shift[C]// Cambridge,USA:International Conference on International Conference on Machine Learning,2015: 448-456.
[9] 李松,魏中浩,张冰尘,等.深度卷积神经网络在迁移学习模式下的SAR目标识别[J].中国科学院大学学报, 2018,35(01): 75-83.
[10] Cherkassay V.The Nature Of Statistical Learning Theory[J].IEEE Transactions on Neural Networks, 2002,38(04): 409-409.
[11] Zeiler M D, Fergus R. Visualizing and Understanding Convolutional Networks[J]. 2013.
[12] Hunt M A, Karnowski T P, Kiest C, et al. Optimizing automatic defect classification feature and classifier performance for post[J]. 2000:116-123.
[13] Jalal A, Kamal S, Kim D. A depth video sensor-based life-logging human activity recognition system for elderly care in smart indoor environments[J]. Sensors, 2014,14(7): 11735-11759.
[14] Islam M A, Akhter S, Mursalin T E, et al. A suitable neural network to detect textile defects[C]//Springer, Berlin,Heidelberg,International Conference on Neural Information Processing, 2006: 430-438.

相似文献/References:

备注/Memo

备注/Memo:
收稿日期:2019-01-11
基金项目:贵州省工业攻关重点项目( [2017]2311)
作者简介:苏杨(1983),男,高级工程师,硕士研究生,研究方向为信息化发展。
更新日期/Last Update: 1900-01-01