MODELING AND OPTIMIZATION OF STEAM DRAG DISTILLATION PROCESS USING ARTIFICIAL INTELLIGENCE AND SIMULATION

Abstract

Introduction:
Distillation by steam dragging is a widely used technique for extracting essential oils, but it presents limitations in thermal efficiency and energy losses. This study applies artificial intelligence (AI) and dynamic simulation to optimize the recovery of essential oil from Ocotea quixos.
Objective:
To develop a predictive model based on artificial intelligence, complemented by dynamic simulations, focused on the process efficiency and the reduction of energy losses.
Materials and Methods:
Automatic learning algorithms were trained using data from a semicontinuous extraction system, calibrating the relationship between steam flow rate, leaf load, and efficiency. The dynamic simulation assessed the thermal evolution of the system through mass and energy balances.
Results and Discussion:
The models enabled operational parameter adjustments, reducing condensate losses by 20% and increasing oil yield by 12%. The simulation revealed improvements in thermal distribution and the design of the rotating drum.
Conclusions:
The combination of AI and simulation substantially enhances the efficiency and sustainability of the extraction process. The proposed approach is replicable to other agro-industrial systems.