arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2103.07303cs.LGcs.NE
收藏

故障检测的二阶分量分析

Second-Order Component Analysis for Fault Detection

East China University of Science and Technology(华东理工大学)

Peng Jingchao, Zhao Haitao, Hu Zhengwei

摘要:

基于神经网络的过程监控越来越受到重视。与经典神经网络相比,高阶神经网络在处理异方差数据方面具有天然的优势。然而,高阶神经网络可能会带来过拟合的风险,并从原始数据和噪声或异常中学习关键信息。正交约束可以大大降低提取特征之间的相关性,从而降低过拟合风险。本文提出了一种新的故障检测方法称为二阶分量分析(SCA)。SCA通过优化正交约束的二阶自编码器排除了过程数据的异方差性。为了解决这一约束优化问题,本文采用了一种几何共轭梯度算法,对Stiefel流形和Euclidean流形的组合进行几何优化。在Tennessee-Eastman基准处理上的大量实验表明,SCA在漏检率(MDR)和虚警率(FAR)方面优于PCA、KPCA和自动编码器。

英文摘要:

Process monitoring based on neural networks is getting more and more attention. Compared with classical neural networks, high-order neural networks have natural advantages in dealing with heteroscedastic data. However, high-order neural networks might bring the risk of overfitting and learning both the key information from original data and noises or anomalies. Orthogonal constraints can greatly reduce correlations between extracted features, thereby reducing the overfitting risk. This paper proposes a novel fault detection method called second-order component analysis (SCA). SCA rules out the heteroscedasticity of pro-cess data by optimizing a second-order autoencoder with orthogonal constraints. In order to deal with this constrained optimization problem, a geometric conjugate gradient algorithm is adopted in this paper, which performs geometric optimization on the combination of Stiefel manifold and Euclidean manifold. Extensive experiments on the Tennessee-Eastman benchmark pro-cess show that SCA outperforms PCA, KPCA, and autoencoder in missed detection rate (MDR) and false alarm rate (FAR).

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑