A Hybrid Modeling Approach for Catalyst Monitoring and Lifetime Prediction,ACS Engineering Au

当前位置： X-MOL 学术 › ACS Eng. Au › 论文详情

Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)

A Hybrid Modeling Approach for Catalyst Monitoring and Lifetime Prediction
ACS Engineering Au ( IF 4.3 ) Pub Date : 2021-09-15 , DOI: 10.1021/acsengineeringau.1c00015
Linh Bui ₁ , Mark Joswiak ₂ , Ivan Castillo ₁ , Ailene Phillips ₃ , Jin Yang ₄ , Daniel Hickman ₅

Affiliation

In this work, we present a hybrid fundamental-empirical model to monitor and predict the catalyst lifetime of an operating industrial reactor. The hybrid model combines a fundamental adiabatic reactor model to calculate the activity of the catalyst bed with an empirical partial least-squares model to predict the catalyst activity at different operating conditions. A baseline model was trained on process data and validated separately using analytical data, showing good agreement between the measured reactant breakthrough of the reactor train and the predicted values from the model over 18 years of continuous operation of four industrial production reactors. To implement the model for catalyst activity monitoring, the model must closely match the current catalyst charge performance. Therefore, the baseline model parameters are updated automatically with new plant data using a filter algorithm.

中文翻译：

催化剂监测和寿命预测的混合建模方法

在这项工作中，我们提出了一个混合的基本经验模型来监测和预测正在运行的工业反应器的催化剂寿命。混合模型结合了计算催化剂床活性的基本绝热反应器模型和用于预测不同操作条件下催化剂活性的经验偏最小二乘模型。基线模型在过程数据上进行了训练，并使用分析数据单独进行了验证，表明反应器系列的测量反应物突破与四个工业生产反应器连续运行 18 年的模型预测值之间具有良好的一致性。为了实施催化剂活性监测模型，该模型必须与当前的催化剂充注性能密切匹配。所以，

更新日期：2021-09-15

点击分享查看原文

点击收藏

公开下载

阅读更多本刊最新论文