AUTHOR=Fajardo Muñoz Sandra E. , Freire Castro Anthony J. , Mejía Garzón Michael I. , Páez Fajardo Galo J. , Páez Gracia Galo J. 

TITLE=Artificial intelligence models for yield efficiency optimization, prediction, and production scalability of essential oil extraction processes from citrus fruit exocarps

JOURNAL=Frontiers in Chemical Engineering

VOLUME=Volume 4 - 2022

YEAR=2023

URL=https://www.frontiersin.org/journals/chemical-engineering/articles/10.3389/fceng.2022.1055744

DOI=10.3389/fceng.2022.1055744

ISSN=2673-2718

ABSTRACT=Excessive demand, environmental problems, and shortages in market-leader countries have drifted the citrus oil (essential oil) market price to unprecedented high levels with negative implications spreading over citrus-oil-dependent secondary industries. However, the high price conditions promote market incentives for the incorporation of new small-scale suppliers as a short-term supply solution for the market. Essential oil chemical extraction via steam distillation stands out as a valuable option for these newcomer suppliers at a lab and small-scale production level. Yet mass-scaling the production demands predicting tools for better controlling the outputs at a large scale. This work provides an intelligent model based on a multi-layer perceptron (MLP) artificial neural network (ANN) for developing a highly-reliable numerical dependency between the chemical extraction output from essential oil steam distillation processes (output vector) and the orange peel mass loading (input vector). From a data pool of twenty-five extraction experiments, fourteen output-input pairs are the training set, six are the testing set, and five cross-compare the model accuracy with traditional numerical approaches. After varying the number of nodes in the hidden layer, a 1-9-1 MLP topology best optimizes statistical parameters (coefficient of determination (R2), and mean square error) of the testing set, achieving a precision near 97.6 %. Our model can to capture the non-linearity behavior when scaling-up production output for mass production processes. Thus we provide a viable route to face the scalability issue with a state-of-the-art computational tool for planning, managing, and mass production of citrus essential oil.