Browse by author
Lookup NU author(s): Professor Anh Phan
Full text for this publication is not currently held within this repository. Alternative links are provided below where available.
© 2026 Elsevier Ltd.Concerns over climate change and fossil fuel dependency have intensified the development of renewable energy and sustainable energy globally. However, conventional mechanistic models often struggle to capture the complex nonlinear behavior of integrated gasification and turbine systems, limiting their predictive accuracy and suitability for real-time applications. This study presents a hybrid modeling framework that integrates a two-stage biomass gasification and gas turbine model with artificial neural networks (ANNs) to enhance prediction accuracy for integrated biomass-to-power performance. More than 5,000 simulation cases were generated using a parametric grid sampling strategy and validated against experimental data from the literature to ensure model fidelity. Spearman’s rank correlation analysis identified the air flow rate and the carbon dioxide (CO₂)-to-fuel ratio as the most influential parameters. The optimized ANN models achieved coefficients of determination (R²) exceeding 0.99 and root mean square errors (RMSE) below 5 %, substantially outperforming conventional modeling approaches. The proposed hybrid ANN–Aspen Plus framework enables rapid and reliable performance prediction, facilitates real-time optimization, and promotes carbon dioxide (CO₂) utilization in advanced biomass-to-power systems.
Author(s): Chandran R, Arpornwichanop A, Mankasem J, Duong LT, Phan AN, Prasertcharoensuk P
Publication type: Article
Publication status: Published
Journal: Journal of Environmental Chemical Engineering
Year: 2026
Volume: 14
Issue: 2
Print publication date: 01/04/2026
Online publication date: 09/02/2026
Acceptance date: 08/02/2026
ISSN (print): 2213-2929
ISSN (electronic): 2213-3437
Publisher: Elsevier Ltd
URL: https://doi.org/10.1016/j.jece.2026.121712
DOI: 10.1016/j.jece.2026.121712
Altmetrics provided by Altmetric
