An Improved Model for Gas-Liquid Flow Pattern Prediction Based on Machine Learning

On the life and habits of gas-core slugs: characterisation of an intermittent horizontal two-phase flow

“…Overall, the model coincided with the algorithm classification in 2630 structures, which corresponds to a 93% match. The quality of the prediction is greater than the predictive accuracy reported using the most recently developed semi-analytical algorithms (around 80%) and is within the same range as those using machine learning techniques (90-96%) [42].…”

Using image‐based machine learning and numerical simulation to predict pesticide inline mixing uniformity

“…The application of ML in agricultural engineering includes not only classification problems, but also regression problems 28,29 . However, the applications of ML algorithms in predicting the inline mixing uniformity of DNIS are still rare, and those application references in chemical industries to realize gas–liquid and liquid–solid flow discrimination are also incompetent for supporting the evaluation of PIMU 30 …”

On the nature of flow patterns and pressure drop fluctuations during flow boiling

“…In addition, several studies have also investigated the behavior of flow pattern and pressure drop through different methods, including image correlation analysis (Rafałko et al (2020)), Green's function (Eyal and Goldstein (2019)), PDF gamma function (Schembri and Bucolo (2011)), Wigner-Ville and Choi-Williams distributions (Du et al (2012)), Adaptive optimal Kernel time-frequency (Zhai et al (2015)), image analysis (Chen et al (2012), Guo et al (2018)) or machine learning (Mask and Wu (2019)). The non-linear dynamics of flow patterns and pressure drop fluctuations were also analyzed in other studies (Rysak et al (2016), Grzybowski and Mosdorf (2014), Grzybowski and Mosdorf (2018), Rahman and Singh (2018), Oliveira et al (2019)).…”