[1]马海娟,杨波,杨思琪,等.基于改进VAE的传感器异常数据检测方法研究[J].计算机技术与发展,2024,34(08):122-127.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0142]
　MA Hai-juan,YANG Bo,YANG Si-qi,et al.Research on Detection Method of Sensor Anomaly Data Based on Improved VAE[J].,2024,34(08):122-127.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0142]
点击复制

基于改进VAE的传感器异常数据检测方法研究

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
34
期数:
2024年08期
页码:
122-127
栏目:
人工智能
出版日期:
2024-08-10

文章信息/Info

Title:
Research on Detection Method of Sensor Anomaly Data Based on Improved VAE
文章编号:
1673-629X(2024)08-0122-06
作者:
马海娟; 杨波; 杨思琪; 杨鑫; 吕沁锐
重庆科技大学 电气工程学院,重庆 401331
Author(s):
MA Hai-juan; YANG Bo; YANG Si-qi; YANG Xin; LYU Qin-rui
School of Electrical Engineering,Chongqing University of Science and Technology,Chongqing 401331,China
关键词:
传感器; 时间序列; 异常检测; 变分自编码器; 动态阈值
Keywords:
sensor; time series; anomaly detection; variational autoencoder (VAE); dynamic threshold
分类号:
TP311.1
DOI:
10.20165/j.cnki.ISSN1673-629X.2024.0142
摘要:
气体传感器在采样过程中受复杂工业环境影响常常产生异常时间序列数据。 传统的时间序列异常检测采用模型预测方法,但没有考虑到时间序列数据的不平衡问题。 因此,提出一种基于改进 VAE 模型的检测方法。 首先,将大量正常时序数据与较少且难以标记的异常时序数据进行合并构建成一个不平衡数据集。 其次,在传统 VAE 模型的基础上采用无监督学习方式,在异常检测分类环节引入动态阈值方法增强网络模型的自适应异常检测能力。 最后,提出一种时序异常检测的组合损失函数,通过集成交叉熵损失函数和 KL 散度进一步提升网络参数优化性能。 实验结果表明,该方法在精确率、召回率以及 F1 值等异常检测性能指标上,比原有的方法有所提升。 该方法在传感器异常数据检测中有着较好的应用。
Abstract:
Gas sensors often produce abnormal time series data due to complex industrial environment during sampling. The traditional a-nomaly detection of time series mainly adopts the model prediction method,but does not consider the imbalance of time series data.Therefore,a detection method based on improved VAE model is proposed.Firstly,a large amount of normal time series data is combined with a small amount of abnormal time series data which is difficult to label to build an unbalanced data set. Secondly,based on the traditional VAE model,unsupervised learning is adopted to enhance the adaptive anomaly detection ability of the network model by intro-ducing dynamic threshold method in anomaly detection classification. Finally,a combined loss function for timing anomaly detection is proposed to further improve the performance of network parameter optimization by integrating cross - entropy loss function and KL divergence. The experimental results show that the proposed method is better than the original method in terms of the accuracy rate,recall rate and F1 value. It has a good application in sensor anomaly data detection.

相似文献/References:

[1]刘记.教育服务网格监控系统的研究与设计[J].计算机技术与发展,2010,(05):67.
　LIU Ji.Research and Design of Monitoring System about Education Service Grid[J].,2010,(08):67.
[2]杜鹏雷 吴晓 杨丽平 江涌.面向精准农业的感知节点传感器驱动与控制[J].计算机技术与发展,2010,(06):233.
　DU Peng-lei,WU Xiao,YANG Li-ping,et al.Drive and Control of Sensor Node Facing Precision Agriculture[J].,2010,(08):233.
[3]赵伟 梁循.互联网金融信息量与收益率波动关联研究[J].计算机技术与发展,2009,(12):1.
　ZHAO Wei,LIANG Xun.Research on Relationship Between Internet Financial Information and Fluctuation of Price - Earnings[J].,2009,(08):1.
[4]詹家宾 卢颖 闫梁.数字式电动操纵负荷系统的研究[J].计算机技术与发展,2009,(12):162.
　ZHAN Jia-bin,LU Ying,YAN Liang.Research of the Digital Electric Control Loading System[J].,2009,(08):162.
[5]何振兴 熊健民 刘幺和 宋庭新.基于WSN的实时监控系统的研究[J].计算机技术与发展,2009,(12):229.
　HE Zhen-xing,XIONG Jian-min,LIU Yao-he,et al.Research of Remote Real - Time Monitoring System Based on WSN[J].,2009,(08):229.
[6]张虹 赵兵 钟华.基于小波多尺度的网络论坛话题热度趋势预测[J].计算机技术与发展,2009,(04):76.
　ZHANG Hong,ZHAO Bing,ZHONG Hua.Hot Trend Prediction of Network Forum Topic Based on Wavelet Multi - Resolution Analysis[J].,2009,(08):76.
[7]常毅 王加阳.时态信息系统转换方法研究[J].计算机技术与发展,2008,(07):15.
　CHANG Yi,WANG Jia-yang.Research on Method of Translating TIS to IS[J].,2008,(08):15.
[8]施尧 赵勇 杨雪洁 赵妹 张燕平 关有训 王克强.基于覆盖算法的大气质量预测[J].计算机技术与发展,2008,(07):190.
　SHI Yao,ZHAO Yong,YANG Xue-jie,et al.Application of Covering Algorithm to Prediction of Air Quality[J].,2008,(08):190.
[9]何星星 孙德山.模糊神经网络与SARIMA结合的时间序列预测模型[J].计算机技术与发展,2008,(08):61.
　HE Xing-xing,SUN De-shan.A Time Series Forecasting Model Using a Hybrid Fuzzy Neural Network and SARIMA[J].,2008,(08):61.
[10]李炜 张义超 卢英 吴建国.基于GPRS环境与安全监测终端设计与实现[J].计算机技术与发展,2008,(09):232.
　LI Wei,ZHANG Yi-chao,LU Ying,et al.Design and Implementation of Environment and Security Terminal Based on GPRS[J].,2008,(08):232.

更新日期/Last Update: 2024-08-10