基于超球体高斯分布的悬浮系统异常检测
Anomaly Detection for Suspension Systems Based on the Gaussian Distribution of Hyperspheres
- 机车电传动 2021年0卷第6期 页码：9-17
- 作者机构：
1.中国空气动力研究与发展中心 设备设计与测试技术研究所，四川 绵阳 621000
2.国防科技大学 智能科学学院，湖南 长沙 410073
- 作者简介：
龙志强（1967—），男，博士，研究员，主要研究方向为悬浮控制技术和故障诊断与容错；E-mail：zhqlong@nudt.edu.cn
- 基金信息：国家“十三五”重点项目课题(2016YFB1200600);国家自然科学基金项目(61801479;62003049)
DOI：10.13890/j.issn.1000-128x.2021.06.002
中图分类号：
扫 描 看 全 文
王平, 梅子, 龙志强. 基于超球体高斯分布的悬浮系统异常检测[J]. 机车电传动, 2021,0(6):9-17.
Ping WANG, Zi MEI, Zhiqiang LONG. Anomaly Detection for Suspension Systems Based on the Gaussian Distribution of Hyperspheres[J]. Electric Drive for Locomotives, 2021,0(6):9-17.
王平, 梅子, 龙志强. 基于超球体高斯分布的悬浮系统异常检测[J]. 机车电传动, 2021,0(6):9-17. DOI： 10.13890/j.issn.1000-128x.2021.06.002.
Ping WANG, Zi MEI, Zhiqiang LONG. Anomaly Detection for Suspension Systems Based on the Gaussian Distribution of Hyperspheres[J]. Electric Drive for Locomotives, 2021,0(6):9-17. DOI： 10.13890/j.issn.1000-128x.2021.06.002.
摘要
悬浮系统的异常指在中低速磁浮列车的实际运行中，悬浮系统工作状态与期望状态不匹配，但是系统没有发生故障。准确预测悬浮系统的异常有助于合理地分配有限的监控资源、提前安排预防性维护计划、减少计划外维修成本、降低事故率。虽然根据《中低速磁浮交通车辆悬浮控制系统技术条件》（CJ/T 458—2014）可获得基于悬浮间隙的经验阈值，但在工程应用中，该方法受到额定悬浮间隙不唯一和外界扰动的影响，常会产生漏报问题，同时悬浮间隙数据的平衡问题增加了异常检测的难度。因此，本文提出了一种基于超球体高斯分布的悬浮系统异常检测方法。该方法首先利用快速沃尔什变换（Fast Walsh-Hadamard Transform，FWHT）技术提取样本特征，并使特征分布空间近似超球体；然后根据样本特征与球心的欧氏距离确定异常阈值。现场试验数据结果表明，与基于经验阈值、基于主成分分析（Principal Component Analysis，PCA）和基于支持向量数据描述（Support Vector Data Description，SVDD）的方法相比，提出的方法在系统异常检测能力方面更具有优越性。
Abstract
The abnormality of the suspension system means that the operating state of the suspension system does not match the expected state, and the system has no faults in the actual operation of the middle-low speed maglev train. Accurately detecting the abnormality of the suspension system helps managers allocate limited monitoring resources reasonably, arrange preventive maintenance plans in advance, reduce unplanned maintenance costs, and reduce accident rates. Although an empirical threshold based on the suspension gap can be obtained according to the "Technical Conditions for the Suspension Control System of Middle-low Speed Maglev Trains CJ/T458—2014", it is affected by the non-unique rated suspension gap and external disturbances, which will cause false negatives in engineering applications. Meanwhile, the problem of the balance of suspension gap data increases the dif ficulty of anomaly detection. Therefore, an anomaly detection method for suspension systems based on the Gaussian distribution of hyperspheres was proposed. Firstly, FWHT(Fast Walsh-Hadamard Transform) technology was used to extract sample features, and the feature distribution space was made approxim ate to a hypersphere. Then, the abnormal threshold was determined according to the Euclidean distance between the sample feature and the center of the sphere. The experimental results of the field data show that, compared with the methods based on empirical threshold, the principal component analysis (PCA) and support vector data description (SVDD), the proposed method is more superior in anomaly detection ability.
关键词
磁浮列车悬浮系统悬浮间隙异常检测FWHTPCASVDD轨道不平顺
Keywords
maglev trainsuspension systemsuspension gapanomaly detectionFWHTPCASVDDtrack irregularity
references
WANG G P, YANG J X, LI R. Imbalanced SVM-based anomaly detection algorithm for imbalanced training datasets[J]. ETRI Journal, 2017, 39(5): 621-631.
KIM M, OU E, LOH P L, et al. RNN-based online anomaly detection in nuclear reactors for highly imbalanced datasets with uncertainty[J]. Nuclear Engineering and Design, 2020, 364: 110699.
LUO M H, WANG K, CAI Z P, et al. Using imbalanced triangle synthetic data for machine learning anomaly detection[J]. Computers, Materials & Continua, 2019, 58(1): 15-26.
CHEN Q, ZHANG A G, HUANG T W, et al. Imbalanced dataset-based echo state networks for anomaly detection[J]. Neural Computing and Applications, 2020, 32(8): 3685-3694.
YOO Y H, KIM U H, KIM J H. Recurrent reconstructive network for sequential anomaly detection[J]. IEEE Transactions on Cybernetics, 2021, 51(3): 1704-1715.
LI L M, YANG R F, GUO C X, et al. The data learning and anomaly detection based on the rudder system testing facility[J]. Measurement, 2020, 152: 107324.
ZHOU X K, HU Y Y, LIANG W, et al. Variational LSTM enhanced anomaly detection for industrial big data[J]. IEEE Transactions on Industrial Informatics, 2021, 17(5): 3469-3477.
WANG X K, WANG H W, WANG Y H. A density weighted fuzzy outlier clustering approach for class imbalanced learning[J]. Neural Computing and Applications, 2020, 32(16): 13035-13049.
CHEN J F, PI D C, WU Z Y, et al. Imbalanced satellite telemetry data anomaly detection model based on Bayesian LSTM[J]. Acta Astronautica, 2021, 180: 232-242.
DU H L, ZHANG Y. Network anomaly detection based on selective ensemble algorithm[J]. The Journal of Supercomputing, 2021, 77(3): 2875-2896.
CHKIRBENE Z, ERBAD A, HAMILA R, et al. Machine learning based cloud computing anomalies detection[J]. IEEE Network, 2020, 34(6): 178-183.
王慧珍, 王立德, 杨岳毅, 等. 基于Logistic集成学习的列车MVB网络异常检测方法研究[J]. 机车电传动, 2021(1): 138-144.
WANG Huizhen, WANG Lide, YANG Yueyi, et al. Anomaly detection for MVB network based on Logistic ensemble learning[J]. Electric Drive for Locomotives, 2021(1): 138-144.
杨欣欣, 李慧波, 胡罡. 一种基于不平衡类深度森林的异常行为检测算法[J]. 中国电子科学研究院学报, 2019, 14(9): 935-942.
YANG Xinxin, LI Huibo, HU Gang. An abnormal behavior detection algorithm based on imbalanced deep forest[J]. Journal of China Academy of Electronics and Information Technology, 2019, 14(9): 935-942.
胡姣姣, 王晓峰, 张萌, 等. 基于深度学习的时间序列数据异常检测方法[J]. 信息与控制, 2019, 48(1): 1-8.
HU Jiaojiao, WANG Xiaofeng, ZHANG Meng, et al. Time-series data anomaly detection method based on deep learning[J]. Information and Control, 2019, 48(1): 1-8.
韩昭蓉, 黄廷磊, 任文娟, 等. 基于Bi-LSTM模型的轨迹异常点检测算法[J]. 雷达学报, 2019, 8(1): 36-43.
HAN Zhaorong, HUANG Tinglei, REN Wenjuan, et al. Trajectory outlier detection algorithm based on Bi-LSTM model[J]. Journal of Radars, 2019, 8(1): 36-43.
姚宇, 冯健, 张化光, 等. 一种基于椭球体支持向量描述的异常检测方法[J]. 山东大学学报(工学版), 2017, 47(5): 195-202.
YAO Yu, FENG Jian, ZHANG Huaguang, et al. Weighted hyper-ellipsoidal support vector data description with negative samples for outlier detection[J]. Journal of Shandong University(Engineering Science), 2017, 47(5): 195-202.
马宏陆, 葛琳琳, 牛强, 等. 一种基于改进EMD的风机振动信号异常检测方法[J]. 南京师大学报(自然科学版), 2017, 40(1): 55-64.
MA Honglu, GE Linlin, NIU Qiang, et al. An improved approach for vibration signal anomaly detection of ventilator based on EMD[J]. Journal of Nanjing Normal University(Natural Science Edition), 2017, 40(1): 55-64.
连鸿飞, 张浩, 郭文忠. 一种数据增强与混合神经网络的异常流量检测[J]. 小型微型计算机系统, 2020, 41(4): 786-793.
LIAN Hongfei, ZHANG Hao, GUO Wenzhong. Netflow anomaly detection based on data enhancement and hybrid neural network[J]. Journal of Chinese Computer Systems, 2020, 41(4): 786-793.
胡文娟. 人工智能的不平衡数据集异常点抽样算法[J]. 计算机仿真, 2020, 37(11): 324-328.
HU Wenjuan. Algorithm for sampling outliers in imbalanced data sets of arti ficial intelligence[J]. Computer Simulation, 2020, 37(11): 324-328.
田增山, 江宇航. 基于快速沃尔什—哈达码变换的WCDMA卫星信号检测算法[J]. 科学技术与工程, 2016, 16(24): 66-70.
TIAN Zengshan, JIANG Yuhang. A WCDMA satellite signal detection algorithm based on fast Walsh-Hadamard transform[J]. Science Technology and Engineering, 2016, 16(24): 66-70.
DUAN L X, XIE M Y, BAI T B, et al. A new support vector data description method for machinery fault diagnosis with unbalanced datasets[J]. Expert Systems with Applications, 2016, 64: 239-246.
DING M M, TIAN H. PCA-based network traf fic anomaly detection[J]. Tsinghua Science and Technology, 2016, 21(5): 500-509.
杨敏, 张焕国, 傅建明, 等. 基于支持向量数据描述的异常检测方法[J]. 计算机工程, 2005, 31(3): 39-42.
YANG Min, ZHANG Huanguo, FU Jianming, et al. Anomaly intrusion detection method based on SVDD[J]. Computer Engineering, 2005, 31(3): 39-42.
0
浏览量
15
下载量
0
CSCD
1
CNKI被引量
- 网络PDF下载
- Meta-XML下载
- BibTeX下载
- Endnote下载
- RefMan 下载
- NoteExpress下载
- NoteFirst下载
关联资源
相关文章
相关作者
相关机构
- 技术支持由北京北大方正电子有限公司提供 湘ICP备20010475号-1
京公网安备11010802024621 - 本系统建议在Chrome、 IE9+ 以上版本浏览器阅读本站内容，360浏览器请切换至极速模式
- Cookies帮助我们提供服务并提供个性化体验。使用本网站，即表示您同意我们使用Cookies