Onel Melis, Kieslich Chris A, Guzman Yannis A, Floudas Christodoulos A, Pistikopoulos Efstratios N
Artie McFerrin Department of Chemical Engineering, Texas A&M University, College Station, TX 77843, USA.
Texas A&M Energy Institute, Texas A&M University, College Station, TX 77843, USA.
Comput Chem Eng. 2018 Jul 12;115:46-63. doi: 10.1016/j.compchemeng.2018.03.025. Epub 2018 Mar 28.
This paper presents a novel data-driven framework for process monitoring in batch processes, a critical task in industry to attain a safe operability and minimize loss of productivity and profit. We exploit high dimensional process data with nonlinear Support Vector Machine-based feature selection algorithm, where we aim to retrieve the most informative process measurements for accurate and simultaneous fault detection and diagnosis. The proposed framework is applied to an extensive benchmark dataset which includes process data describing 22,200 batches with 15 faults. We train fault and time-specific models on the prealigned batch data trajectories via three distinct time horizon approaches: one-step rolling, two-step rolling, and evolving which varies the amount of data incorporation during modeling. The results show that two-step rolling and evolving time horizon approaches perform superior to the other. Regardless of the approach, proposed framework provides a promising decision support tool for online simultaneous fault detection and diagnosis for batch processes.
本文提出了一种用于间歇过程过程监控的新型数据驱动框架，这是工业中实现安全可操作性并最大限度减少生产率和利润损失的一项关键任务。我们利用基于非线性支持向量机的特征选择算法处理高维过程数据，旨在获取最具信息性的过程测量值，以进行准确且同步的故障检测与诊断。所提出的框架应用于一个广泛的基准数据集，该数据集包含描述22200个批次且带有15种故障的过程数据。我们通过三种不同的时间范围方法，在预先对齐的批次数据轨迹上训练故障和特定时间模型：一步滚动、两步滚动和逐步演变，后者在建模过程中会改变数据纳入量。结果表明，两步滚动和逐步演变的时间范围方法表现优于其他方法。无论采用哪种方法，所提出的框架都为间歇过程的在线同步故障检测与诊断提供了一个有前景的决策支持工具。
