Real-time Monitoring of Two-phase Flow Using Electrical Impedance Tomography with Convolutional Neural Network

Two-phase flow is common in industrial systems, and monitoring the state of two-phase flow is important for production and manufacturing. Electrical impedance tomography (EIT) is a promising non-invasive visualization technique for monitoring two-phase flows. However, in stratified flow, traditional algorithms have low imaging accuracy and can produce artifacts, making it difficult to accurately determine the interface between the two phases. In this paper, we propose a real-time monitoring method for stratified two-phase flow using EIT with a convolutional neural network (CNN). We use the Impedance Matrix (IM) to train the CNN model to reconstruct stratified two-phase flow images. The proposed method is evaluated on a series of simulated data sets with different noise levels and compared with traditional image reconstruction methods such as the NOSER algorithm. Furthermore, the effectiveness of the method for real-time monitoring of multiphase flows in pipes is verified. The results show that the proposed CNN-based method is superior to traditional methods in terms of imaging reconstruction accuracy and noise robustness and can be used for real-time monitoring of stratified two-phase flow. This study provides a promising method for real-time monitoring of stratified two-phase flow.