Continuously operated pharmaceutical manufacturing lines often consist of a wet granulation unit operation, followed by a (semi-) continuous dryer. The operating conditions of the dryer are crucial for obtaining a desired final granule moisture. Commercially available dryers lack of a thorough online measurement of granule moisture during the drying process. However, this information could improve the operation of the equipment considerably, yielding a granule moisture close to the desired value (e.g., by drying time and process parameter adjustments in real-time). The paper at hand proposes a process model, which can be parameterized from a very limited number of experiments and then be used as a so-called soft sensor for predicting granule moisture. It utilizes available process measurements for the estimation of the granule moisture. The development of the model as well as parameter identification and validation experiments are provided. The proposed model paves the way for the application of sophisticated observer concepts. Possible future activities on that topic are outlined in the paper.

中文翻译：

---

ConsiGma™-25分段流化床干燥机中颗粒水分的终点预测。

连续操作的制药生产线通常包括湿法制粒单元操作，然后是（半）连续干燥机。干燥机的操作条件对于获得所需的最终颗粒水分至关重要。市售的干燥机在干燥过程中缺乏对颗粒水分的全面在线测量。但是，此信息可以显着改善设备的操作，产生接近所需值的颗粒水分（例如，通过干燥时间和实时过程参数调整）。这篇论文提出了一种过程模型，可以通过很少数量的实验对其进行参数设置，然后将其用作预测颗粒水分的所谓软传感器。它利用可用的过程测量来估算颗粒的水分。提供了模型的开发以及参数识别和验证实验。所提出的模型为复杂的观察者概念的应用铺平了道路。本文概述了将来可能在该主题上开展的活动。