文献类型：期刊文献

英文题名：One-Class Classification Constraint in Reconstruction Networks for Multivariate Time Series Anomaly Detection

作者：Li, Jiazhen[1];Yu, Zhenhua[1];Jiang, Qingchao[1];Cao, Zhixing[1]

机构：[1]East China Univ Sci & Technol, Key Lab Smart Mfg Energy Chem Proc, Minist Educ, Shanghai 200237, Peoples R China

年份：2025

卷号：74

外文期刊名：IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT

收录：;EI(收录号：20251118042249);WOS:【SCI-EXPANDED(收录号:WOS:001453415000046)】；

基金：This work was supported in part by the National Natural Science Foundation of China under Grant 62322309, in part by Shanghai Pilot Program for Basic Research under Grant 22TQ1400100-16, and in part by Shanghai Science and Technology Innovation Action Plan under Grant 23S41900500.

语种：英文

外文关键词：Anomaly detection; Data models; Feature extraction; Time series analysis; Convolution; Attention mechanisms; Transformers; Autoencoders; Training; Stability analysis; deep learning; multivariate time series (MTS); one-class classification; transformer

摘要：Detecting the anomalies in multivariate time series (MTS) data is crucial for maintaining the stability of industrial manufacturing processes and biochemical operations. However, current methods often focus on capturing the normal patterns of training data while overlooking the potential of latent representations. This research introduces the hypersphere constraint network (HSC), an innovative self-supervised model for anomaly detection in MTS. This approach uniquely integrates a one-class classification framework to regulate latent distribution. First, the HSC employs a temporal convolutional network (TCN) and a multilayer perceptron (MLP) to extract latent representations of input data, imposing constraints on the latent distribution to achieve a one-class loss. Second, a self-attention mechanism is applied to reconstruct the input data and calculate the reconstruction loss. Anomalies are identified by combining the one-class loss with the reconstruction loss. By integrating one-class classification with a reconstruction-based method, the HSC significantly increases sensitivity to anomalous data, enhancing the distinction between normal and abnormal data. Evaluations on three real-world datasets and a simulated dataset demonstrate that the HSC model outperforms existing state-of-the-art methods in anomaly detection.