A classification AI model to predict choking of vibrating screen based on DEM and machine learning

Screening is a complicated process for classifying granular materials according to size. Choking is a vital issue in screening. It may occur when the particle flow along a screen is too slow, but slow particle flow and long residence time are beneficial to sieving performance. Therefore, a model to judge whether choking happens is useful for finding optimal operating conditions. Here, a classification model to predict screen choking is proposed by combining DEM simulation and machine learning. The model can consider various key controlling variables for particle properties and operating conditions. Using the model, safe operation condition regions without choking can be identified. Then, combining the model with our previous machine learning based process model, we can design a screening process with the desired performance. The work also shows a way of using machine learning to predict critical phenomena in particle flow. • Classification AI model for screen chocking developed based on DEM data. • Model can be used to obtain safe operating conditions for screening. • Process optimization by combining classification and predictive AI models • Model predicts feeding threshold for screen chocking. • Classification AI model for critical phenomena of granular matter