Open AccessModelling and Prediction of Compressive Strength of Hydraulic Concrete Structure in Multiproduct Batch Plant Design for Protein Production using Extreme Gradient Boosting Regressor with Grid Search support
Author(s) -
Youness El Hamzaoui,
Juan Antonio,
A. Arellano,
Youness El
Publication year - 2021
Publication title -
journal of pharmaceutical research and development
Language(s) - English
Resource type - Journals
ISSN - 2694-5614
DOI - 10.47485/2694-5614.1011
Subject(s) - mathematical optimization , grid , computer science , hyperparameter optimization , boosting (machine learning) , compressive strength , process (computing) , mathematics , support vector machine , artificial intelligence , materials science , geometry , composite material , operating system
This work deals with the problem of modeling and prediction of compressive strength of concrete structure in multiproduct batch plant design of protein production found in a chemical engineering process with uncertain demand. Modeling the strength of concrete for this process is very complex. However, it can be solved by minimizing the investment cost. Therefore, the aim of this work is to minimize the investment cost and find out the number and size of parallel equipment units in each stage. For this purpose, it is proposed to solve the problem by using extreme gradient boosting regressor with grid search support (XGBoost), could be interpreted as an optimization algorithm on a suitable cost function, which take into account, the uncertainty on the demand using gaussian process modeling. The results about number and size of equipment’s, investment cost, production time, process time and idle times in plant obtained by light gradient boosted trees regressor are the best. This methodology can help the decision makers and constitutes a very promising framework for finding a set of “good solutions”.