Pretreatment Model

Overview

Different data-driven and knowledge-driven models for the dilute acid pretreatment of wheat straw

Employment

The repository contains the following folders:

• design_of_experiments: folder with the files for calculating the designs of expeirments
• easyGSA: empty folder to install the easyGSA toolbox
• experimenta_data: folder containing a .mat file with the experimental data of the dilute acid pretreatment
• identifiability_analysis: folder with the scripts/functions for performing the identifiability analysis (main file: identifiability_analysis.m)
• model_gpr: folder with the script to fit a GPR model to the experimental data
• model_mechanistic: folder with the scripts/functions for the mechanistic model and the validation (main file: sim_pretreatment.m, model_validation.m)
• model_rsm: folder with the script to fit a RSM model to the experimental data
• MOSKopt: empty folder to install MOSKopt
• optimization_gpr: folder with the script to run the optimization problem with the GPR model
• optimization_mechanistic: folder with the scripts/functions to run the optimization problem with the mechanistic model (main file: opt_mechanistic.m)
• optimization_MOSKopt: folder with the scripts/functions to run the optimization problem with the mechanistic model and the MOSKopt solver (main file: rs_MOSKopt.m)
• parameter_estimation_initial: folder with the scripts/functions to run the intial parameter estimation before the identifiability analysis (main file: parameter_estimation_initial.m)
• parameter_estimation_main: folder with the scripts/functions to run the main parameter estimation after the identifiability analysis (main file: parameter_estimation_main.m)
• sensitivity_analysis: folder with the scripts/functions to run the sensitivity analysis with the easyGSA toolbox (main file: sensitivity_analysis.m)
• uncertainty_analysis: folder with the script to run the uncertainty analysis

Prerequisites

• MATLAB 2020b or higher (https://www.mathworks.com/)
• R 4.0.5 or higher (https://www.r-project.org/)

The frameworke furthermore utilizes the easyGSA toolbox and the MOSKopt solver developed by Resul Al (resal@kt.dtu.dk). They are available on the GSI-Lab GitHub page:

• easyGSA (https://github.com/gsi-lab/easyGSA)
• MOSKopt (https://github.com/gsi-lab/MOSKopt)

Developer

Nikolaus Vollmer (nikov@kt.dtu.dk) - PROSYS Research Center, Department of Chemical and Biochemical Engineering, Technical University of Denmark

Acknowledgements

This work is part of the Fermentation-Based Biomanufacturing Initiative (http://www.fbm.dtu.dk) at the Technical University of Denmark and received funding by the Novo Nordisk Foundation (Grant no. NNF17SA0031362)

Parts of the code stem from:

• Sin et al., 2010, Assessing reliability of cellulose hydrolysis models to support biofuel process design—Identifiability and uncertainty analysis, Computers & Chemical Engineering, Vol. 34 (9), pp. 1385-1392. https://doi.org/10.1016/j.compchemeng.2010.02.012
• Sin and Gernaey, 2016, Data Handling and Parameter Estimation, in Experimental methods in wastewater treatment, pp. 201-234, ISBN: 9781780404745.