A Unifying and Integrated Framework for Feature Oriented Analysis of Batch Processes

Abstract

We present a data analytics framework for offline analysis of batch processes. The framework provides a unified setting for implementing several variants of feature oriented analysis proposed in the literature, including a new methodology based on the process variables’ profiles presented in this article. It also integrates feature generation and feature analysis, in order to speed up the data exploration cycle, which is especially relevant for complex batch processes. The FOBA (Feature Oriented Batch Analytics platform) is described in detail and applied to several case studies regarding different analysis goals: visualization of the differences between the operation of two industrial units (dryers), quality prediction, and end-of-batch process monitoring. The performance of the proposed methodology is also critically assessed and compared with other alternative analytical approaches currently in use.

Predicting ships' CO2emissions using feature‐oriented methods

“…A similar procedure of profile assignment and feature extraction is applied to every other variable, and therefore, PdF transforms the measured data into a feature matrix F PdF . Further details of PdF can be consulted in the original paper …”

“…As stated before, the feature‐oriented methods contemplated in this paper, namely, PdF and SPA, are compared with the standard method based on the AV.…”

“…Further details of PdF can be consulted in the original paper. 22 Here, SPA is a feature extraction procedure based on the variables' statistical moments, and typical features include the mean, variance, covariance between variables, kurtosis, and skewness. Whereas AV computed the average value of the profile obtained for variable j in voyage n (x n, ,1∶K n ), SPA computes additional moments such as the variance of the profile, its kurtosis, and skewness.…”

“…These methods are simpler to apply because they do not require synchronization and tend to be more parsimonious since the number of model parameters is smaller. Examples of feature‐oriented methods include profile‐driven features (PdF) recently proposed by Rendall et al and statistical pattern analysis (SPA) proposed by He and Wang . These techniques compress each variable into a small number of features that can be utilized for data‐driven model building.…”

Monitoring batch processes with dynamic time warping and k-nearest neighbours

“…Adaptations to the DTW algorithm have been proposed [9,10] which aim to improve performance of online-alignment by limiting the degree of permitted re-alignment at each time-point. Other authors have proposed methods which avoid explicit synchronisation of the data, for example by applying feature extraction [11]. In this work, the distance measure provided by DTW is used to quantify an ongoing batch's distance to NOC batches so that the DTW alignment itself is the basis of the monitoring framework.…”