Optimizing the Parameters of Fluidized Bed Dryer

The present study centres on the application of an all-encompassing methodology for optimizing the parameters of Fluidized Bed Dryer (FBD). First, a dataset with various FBD attributes are subjected to Principle Component Analysis (PCA). Subsequently, a variety of regression algorithms, including multiple linear regressors, support vector classifiers, decision trees, random forests, and gradient boosting regressors, are employed to analyze the identified key attributes. These attributes comprise the actual values of inlet air flow, inlet air temperature, exhaust air temperature, loss on drying, product temperature, and total drying time. When it comes to predicting the total drying time, the gradient boosting regressor performs better than others. A novel optimization technique called Bayesian optimization is used to find the optimized values of highly correlated features with the total drying time in order further maximize efficiency. This integrated approach uses both sophisticated machine learning algorithms for predictive modelling and optimization techniques to fine-tune specific parameters that are critical for the Fluidized Bed Dryer. The study's findings provide useful information about enhancing FBDs' overall performance and efficacy, which advances drying processes in a variety of industries.