AI-chemometric assisted real-time monitoring of tryptophan fermentation process using a sensor fusion strategy.
AI summary
75% confidenceThis MES study investigates Fermentation, In-line monitoring, Machine learning. The research focuses on system optimization and characterization.
Generated by MESSAI extraction pipeline · review against source PDF
Reported parameters
No values extracted from this paper yet.
No 3D model is mapped to this paper yet. Parameter ranges above still place reported values on the literature distribution.
Abstract
Tryptophan, an essential amino acid, crucially impacts neuronal function, metabolism, immunity, and gut homeostasis. Microbial fermentation is the mainstream method for tryptophan production. The precise production process is essential for ensuring both high quality and optimal yield. This study aims to utilize AI-chemometrics methods to achieve the integrated monitoring of multi-source sensors. First, machine learning methods were applied to build in-line NIR prediction models. Then, a sensor fusion strategy was introduced to established the multivariate statistical process control (MSPC) model based on five in-line sensor data. The results showed that Gaussian process regression models were best for bacterial optical density, residual sugar, and tryptophan concentration. The validation sets RPD were 5.686, 3.297, and 3.130, respectively. MSPC charts synergistic analysis based on feature-level fusion enables real-time simultaneous detection of multi-source anomalies. This study provides an effective quality control strategy for food fermentation process to ensure consistent, stable and controllable product quality.