Pressure Drop Prediction in Fluidized Dense Phase Pneumatic Conveying using Machine Learning Algorithms

Modeling of pressure drop in fluidized dense phase conveying (FDP) powders is a tough work as the flow comprises various interactions among solid, gas and pipe wall. It difficult to incorporate these into model. The depends on flow, material geometrical parameters. existing models show high error when applied other pipeline configurations varying lengths or diameters. current study investigates capability machine learning (ML) techniques estimate FDP powders. Pneumatic experimental data were used for training network then predicting drop. For estimating drop, four distinct ML algorithms light gradient boosting (LighGBM)), multilayer perception (MLP), K-nearerst neighbors (KNN), extreme (XGBoost), selected. XGBoost model performed better than chosen with ±5% margin while testing data, ±10% validating data. MLP, XGBoost, KNN, LightGBM predicted MAE 5.05, 1.19, 5.72, 2.85, respectively, well Among considered, using algorithm best, whereas KNN poorly