[1]李春生,田梦晴,张可佳.基于 Bi-LSTM 网络的管道异常数据检测方法[J].计算机技术与发展,2023,33(06):215-220.[doi:10. 3969 / j. issn. 1673-629X. 2023. 06. 032]
　LI Chun-sheng,TIAN Meng-qing,ZHANG Ke-jia.Pipeline Anomaly Data Detection Method Based on CNN and Bi-LSTM Network[J].,2023,33(06):215-220.[doi:10. 3969 / j. issn. 1673-629X. 2023. 06. 032]

点击复制

基于 Bi-LSTM 网络的管道异常数据检测方法()

分享到：

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

卷:

33

期数:

2023年06期

页码:

215-220

栏目:

新型计算应用系统

出版日期:

2023-06-10

文章信息/Info

Title:

Pipeline Anomaly Data Detection Method Based on CNN and Bi-LSTM Network

文章编号:

1673-629X(2023)06-0215-06

作者:

李春生; 田梦晴; 张可佳

东北石油大学 计算机与信息技术学院,黑龙江 大庆 163319

Author(s):

LI Chun-sheng; TIAN Meng-qing; ZHANG Ke-jia

School of Computer and Information Technology,Northeast Petroleum University,Daqing 163319,China

关键词:

异常点检测; 管道运行数据; 卷积神经网络; 双向长短期记忆网络; 时序数据

Keywords:

anomaly detection; pipeline operation data; convolutional neural network; bidirectional long short-term memory network; timeseries data

分类号:

TP391

DOI:

10. 3969 / j. issn. 1673-629X. 2023. 06. 032

摘要:

在管道运行过程中,受技术、计量仪器和自然环境等影响,导致管道数据经常出现异常值,影响调度人员无法进行正确的决策,不利于管道监控系统的安全稳定运行。 传统的时序数据异常检测方

法的准确率和检测速度得不到保证。 针对该问题,提出一种基于卷积神经网络( Convolutional Neural Network,CNN) 和双向长短期记忆( Bi-directional Long-ShortTerm Memory,Bi-LSTM) 网络

的管道异常数据检测方法。 首先,研究管道异常数据的表征及异常数据的产生原因,对管道数据进行野点剔除、均值填充和归一化处理,后通过 CNN 对处理后的管道数据进行特征提取;其次,利用 Bi-LSTM 网络充分挖掘管道数据间的规律,训练得到预测模型;再次,确定动态阈值,通过计算预测值与真实值误差并与阈值进行比较,检测异常数据;最后,在真实应用场景测试,通过设计一系列对比实验验证了该方法在处理速度和检测准确率等方面具有明显优势,且检测异常点的准确率高于同类算法。

Abstract:

In the process of pipeline operation,due to the influence of technology,measuring instruments and natural environment,pipelinedata often appear abnormal values,which affects the scheduler cannot make correct decisions and is not conducive to the safe and stableoperation of pipeline monitoring system. The accuracy and detection speed of traditional time series data anomaly detection methods arenot guaranteed. Aiming at this problem,we propose a pipeline abnormal data detection method based on convolutional neural networkand bidirectional long - short - term memory neural network. Firstly, the representation of abnormal pipeline data and the causes ofabnormal data are studied, outfield elimination, mean filling and normalization are carried out on the pipeline data, and then thecharacteristics of pipeline data after processing are extracted by CNN. Secondly,Bi-LSTM network is used to fully mine the law betweenpipeline data and build a prediction model. Thirdly,determine the dynamic threshold,calculate the error between the predicted value andthe true value and compare it with the threshold value to detect abnormal data. Finally,the proposed method is applied in a real world todemonstrate our ultra-short-term working condition prediction method that achieves superior results for prediction accuracy and runningspeed when compared with other methods,and the accuracy of detecting outliers is higher than that of similar algorithms.

常用功能