Critical evaluation of approaches for on-line batch process monitoring

Sprang, E.N.M. van; Ramaker, Henk‐Jan; Westerhuis, Johan A.; Gurden, Stephen P.; Smilde, Age K.

doi:10.1016/s0009-2509(02)00338-x

Cited by 114 publications

(90 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, in four of the nine combinations, false positives appeared in Q. As in other studies [12], it appears that the validation residuals are predisposed to be larger than those in the calibration set, and thus, false positives for Q should be expected especially for such a small number of batches.…”

Section: A Fault Detection and Diagnosissupporting

confidence: 58%

Fault Detection and Diagnosis in an Industrial Fed-Batch Cell Culture Process

Gunther¹,

Conner²,

Seborg³

2007

Biotechnol Progress

View full text Add to dashboard Cite

Abstract-This paper applies multivariate statistical process control (MSPC) techniques to pilot plant fermentation data for the purpose of fault detection and diagnosis. Data from ten batches, nine normal operating conditions (NOC) and one failed, were available. A principal component analysis (PCA) model was constructed from eight NOC batches, while the remaining NOC batch was used for model validation. Subsequently, the model was used to successfully detect (both offline and online) a process abnormality in the failed batch and diagnose the factors contributing to the fault. These monitoring results agree with the observed biological phenomena encountered during this batch.

show abstract

Section: A Fault Detection and Diagnosissupporting

confidence: 58%

Fault Detection and Diagnosis in an Industrial Fed-Batch Cell Culture Process

Gunther¹,

Conner²,

Seborg³

2007

Biotechnol Progress

View full text Add to dashboard Cite

show abstract

“…The SBR and DuPont datasets are also used in the review of Van Sprang et al [78] and the comparison between global, evolving, and local PCA models for monitoring by Ramaker et al [64]. These two papers also included three additional datasets: (i) an industrial multistage polymerization set of 47 normal and 3 abnormal batches (again provided by DuPont) [45], (ii) a collection of 67 normal and 3 faulty runs of an industrial batch polymerization of PVC [74], and (iii) a biochemical conversion set of 27 normal batches and 1 faulty batch [9].…”

Section: Introductionmentioning

confidence: 99%

An extensive reference dataset for fault detection and identification in batch processes

Impe¹,

Gins²

2015

Chemometrics and Intelligent Laboratory Systems

View full text Add to dashboard Cite

Close process monitoring (i.e., detection and identification of disturbances) is important to achieve high process efficiency and safety. The Tennessee Eastman process is an extensive benchmark dataset for fault detection and identification, but it is only representative for continuous processes because it does not contain the inherent nonstationarity that complicates monitoring of batch processes. Nevertheless, batch processes also play an important role in many types of industry. This paper therefore presents an extensive reference dataset for benchmarking data-driven methodologies for fault detection and identification in batch processes. The original Pensim model [10] is expanded with sensor noise. By changing the properties of the initial conditions and/or model parameters, four subsets of different complexity are generated, each containing 400 batches with normal operation. To correctly assess the fault detection and identification in batch processes, 15 faults are simulated with various amplitudes and onset times for a total of 22,200 faulty batches for each subset, or 90,400 batches in total. Analysis of the data indicates that the presented types of process faults and their various amplitudes in each of the four subsets present a suitable benchmark for fault detection and identification in batch processes. The dataset is freely available at http://cit.kuleuven.be/biotec/batchbenchmark.

show abstract

“…Multi-way principal component (MPCA) and multi-way partial least squares (MPLS) methods have laid a foundation for introducing multivariate statistical analysis method to batch processes monitoring [6][7][8]. After that, some researches and improvements for batch processes monitoring have been developed [9][10][11][12][13][14][15][16][17][18]. However, it is noted that there exits an important information on multiplicity in batch processes, which can not be mined by traditional MPCA and MPLS method.…”

Section: Introductionmentioning

confidence: 99%