Monitoring Multimode Nonlinear Dynamic Processes: An Efficient Sparse Dynamic Approach With Continual Learning Ability
Monitoring Multimode Nonlinear Dynamic Processes: An Efficient Sparse Dynamic Approach With Continual Learning Ability
复制标题
监测多模非线性动态过程：具有持续学习能力的高效稀疏动态方法
DOI：
10.1109/tii.2022.3215971
复制
发表时间：
2023-07
影响因子：
12.3
通讯作者：
Xia Hong
中科院分区：
文献类型：
--
作者：
Jingxin Zhang;Maoyin Chen;Xia Hong
工业过程通常在多种模式下运行，全球监测方法建立在针对每种模式的本地模型的基础上，需要来自所有潜在模式的完整数据。然而，实际数据的生成和收集是稳定的，这使得处理变得困难，如果不是不可能的话。本文提出了一种高效的多模非线性动态过程监测稀疏动态内主成分分析算法，旨在建立对连续模式具有持续学习能力的单一监测模型。为了减少存储和计算成本，基于余弦相似度从每种模式中选取少量具有代表性的数据，在新模式到来时进行重放进行再训练，足以反映每种模式的运行情况。受重放连续学习的启发，所有现有模式的数据通过各自的统计进行预处理，然后将其视为一个完整的数据集，然后构建单个多模式监测模型。通过向量自回归模型从原始数据中提取多模态动态潜变量。因此，该方法不受模式相似度的约束，适用于多种模式，便于长期监测任务。此外，该方法可以处理非线性问题，并增加了正则化项以避免潜在的过拟合问题。通过一个连续搅拌罐式加热器和一个实际的工业系统，与现有的多模态监测方法进行了比较，验证了该方法的有效性。
Industrial processes generally operate under multiple modes and a global monitoring approach, built upon combining local models that are aimed at each mode, requires complete data from all potential modes to be available. However, practical data are generated and collected in a steady stream, which makes it difficult if not impossible to process. This article proposes an efficient sparse dynamic inner principal component analysis algorithm for multimode nonlinear dynamic process monitoring, which aims to build a single monitoring model with continual learning ability for successive modes. To reduce the storage and computational costs, only a few representative data from each mode are selected based on cosine similarity and replayed for retraining when a new mode arrives, which are sufficient to reflect the operating condition of each mode. Inspired by replay continual learning, data from all existing modes are preprocessed by their own statistics and then regarded as a whole dataset, followed by building a single multimode monitoring model. The multimode dynamic latent variables are extracted from data in raw format, via a vector autoregressive model. Therefore, the proposed method is not constrained by the mode similarity, which makes it appropriate for diverse modes and convenient for long-term monitoring tasks. Besides, the proposed method can deal with nonlinearity and a regularization term is added to avoid the potential overfitting issue. Compared with state-of-the-art multimode monitoring methods, the effectiveness of the proposed approach is demonstrated by a continuous stirred tank heater and a practical industrial system.
登录
查看更多内容
基于结构字典学习的多模式过程监控及其在铝电解过程中的应用
DOI：
10.1109/tase.2020.2984334
发表时间：
2020
影响因子：
5.6
作者：
Keke Huang;Yiming Wu;Chunhua Yang;Gongzhuang Peng;Weiming Shen
通讯作者：
Weiming Shen
基于混合典型变量分析模型的多模式动态过程监测
影响因子：
4.2
作者：
Qiaojun Wen;Zhiqiang Ge;Zhihuan Song
通讯作者：
Zhihuan Song
DOI：
10.1109/tase.2022.3144288
发表时间：
2023-01
影响因子：
5.6
作者：
Jingxin Zhang;Donghua Zhou;Maoyin Chen;Xia Hong
通讯作者：
Jingxin Zhang;Donghua Zhou;Maoyin Chen;Xia Hong
影响因子：
12.3
作者：
K. Pilario;Yi Cao
通讯作者：
K. Pilario;Yi Cao
自适应协整分析和改进的具有持续学习能力的 RPCA 用于监测多模非平稳过程
DOI：
10.1109/tcyb.2021.3140065
发表时间：
2022-02
期刊：
IEEE TRANSACTIONS ON CYBERNETICS,
影响因子：
--
作者：
Jingxin Zhang;Donghua Zhou;Maoyin Chen
通讯作者：
Maoyin Chen