Dynamic model-based fault diagnosis for (bio)chemical batch processes

摘要

To ensure a constant and satisfactory product quality, close monitoring of batchprocesses is an absolute requirement in the chemical and biochemical industry.Principal Component Analysis (PCA)-based techniques exploit historical databases forfault detection and diagnosis of the current batch run. To handle the dynamic nature ofbatch processes, dedicated techniques such as Batch Dynamic PCA (BDPCA [1]) andAuto-Regressive PCA (ARPCA [2]) have been developed. In this paper, the faultdetection and diagnosis performance of BDPCA and ARPCA is compared with standardmulti-way PCA (MPCA [3]) on an extensive dataset of a penicillin fermentation. ForMPCA, an additional batch-wise normalization improves the detection of faults on thefeed rate and dissolved oxygen (DO). ARPCA clearly outperforms MPCA and BDPCAin detection speed of drifts on the aeration rate, stirrer power and DO. Drifts on the feedrate are detected slightly faster by MPCA and BDPCA. Fault diagnosis performance iscomparable for each model and allows for the correct fault diagnosis in most cases.