Physics-informed recurrent neural networks and hyper-parameter optimization for dynamic process systems

doi:10.1016/j.compchemeng.2023.108195

Physics-informed recurrent neural networks and hyper-parameter optimization for dynamic process systems

2023-05-01

35

PRE

AI

DOI：10.1016/j.compchemeng.2023.108195

Original

Original request for help

Save

Abstract

En 中文

Many of the processes in chemical engineering applications are of dynamic nature. Mechanistic modeling of these processes is challenging due to the complexity and uncertainty. On the other hand, recurrent neural networks are useful to be utilized to model dynamic processes by using the available data. Although these networks can capture the complexities, they might contribute to overfitting and require high-quality and adequate data. In this study, two different physics-informed training approaches are investigated. The first approach is using a multiobjective loss function in the training including the discretized form of the differential equation. The second approach is using a hybrid recurrent neural network cell with embedded physics-informed and data-driven nodes performing Euler discretization. Physics-informed neural networks can improve test performance even though decrease in training performance might be observed. Finally, smaller and more robust architecture are obtained using hyper-parameter optimization when physics-informed training is performed.

Keywords:

Machine learning

Recurrent neural networks

Physics -informed neural networks

Hybrid neural networks

Hyper -parameter optimization

Physics-informed recurrent neural networks and hyper-parameter optimization for dynamic process systems

Abstract

Journal

Organization