A Comprehensive Monitoring Strategy for Multiphase Batch Processes: Dual Consideration Metrics and Bilateral Threshold Linkage Monitoring Framework

Multiphase characteristics are inherent attributes in batch industrial processes. When dealing with multiphase batch processes, not only do the complex variable relationships contained in different phases need to be considered but also some small faults are prone to be masked by transition phases, presenting significant challenges to traditional process monitoring methods. To address these challenges, this article proposes dual consideration metrics and bilateral threshold linkage monitoring (DCMs-BTLM) framework for multiphase batch processes. First, the application of a sliding window methodology facilitates the segmentation of batch data. Subsequently, the dissimilarity analysis is employed to scrutinize fluctuations in the correlation between adjacent windows. Next, the novel DCMs are established by integrating a proposed distance metric with the dissimilarity analysis. Finally, a BTLM framework is proposed to effectively achieve phase division and process monitoring. The DCMs-BTLM strategy adopts a dual-measurement approach, incorporating both distance and correlation metrics in the phase division section, thereby enabling a more rational demarcation of transition phases. Furthermore, the introduced BTLM framework not only effectively pinpoints transition phases of test batches but also ensures a robust fault detection capability, thereby enhancing overall process control. The effectiveness of DCMs-BTLM is verified by the benchmark fed-batch penicillin fermentation process and semiconductor etching process.