A new deep neural network framework with multivariate time series for two-phase flow pattern identification

作者：

Highlights：

• A powerful deep learning algorithm is developed for flow pattern recognition.

• The accuracy of proposed model for flow pattern recognition is 99.3%.

• The proposed model has remarkable generalization ability and robustness.

• This study opens up a new venue for investigating industrial multiphase flow.

摘要

•A powerful deep learning algorithm is developed for flow pattern recognition.•The accuracy of proposed model for flow pattern recognition is 99.3%.•The proposed model has remarkable generalization ability and robustness.•This study opens up a new venue for investigating industrial multiphase flow.

论文关键词：Gas-water two phase flow,Flow pattern identification,Feature extraction,Deep learning classifiers

论文评审过程：Received 23 September 2021, Revised 27 March 2022, Accepted 29 May 2022, Available online 6 June 2022, Version of Record 9 June 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117704