Optimization of photo-fermentation bio-hydrogen production from corncob via genetic algorithm optimized neural network and response surface method model

• Published in: International journal of hydrogen energy 2024-11
• Main Authors: Ai, Fuke, Hu, Yuan, Kang, Kang, Lam, Su Shiung, Foong, Shin Ying, Yong, Cheng, Li, Yameng, Zhang, Quanguo, Zhang, Yang, Zhu, Shengnan, Lv, Xianchao, Cheng, Axing, Zhang, Zhiping
• Format: Article
• Language: English
• Subjects: Corncob; Cumulative hydrogen yield; Enterobacter hormaechei; HAU-M2; Root mean square error
• ISSN: 0360-3199
• DOI: 10.1016/j.ijhydene.2024.10.325

Photosynthetic hydrogen-producing bacteria play a critical role in photo-fermentation bio-hydrogen production (PFHP) and optimizing the operating conditions is essential for improving hydrogen yield. In this study, corncob was used as raw material for fermentation, Enterobacter hormaechei (EH) was combined with the photosynthetic hydrogen-producing bacterial community HAU-M1 to form a new bacterial community, HAU-M2, for hydrogen production via PFHP. Response surface method (RSM) model and a genetic algorithm optimized neural network (GANN) model were used and compared to optimize the operating conditions of the PFHP process. The results showed that the GANN model showed enhanced optimization abilities. Under optimal conditions, the cumulative hydrogen yield was 51.96 mL/g TS. The energy recovery efficiency in the GANN experimental group (3.59 ± 0.12%) increased by 55% compared to the control group (2.31 ± 0.13%). This study provides valuable insights and references for the resource utilization of agricultural waste and the clean production of renewable energy. [Display omitted] •GANN shows superior performance in predicting PFHP process.•RSM model accurately reflects the interaction among various factors.•GANN model shows enhanced optimization capability.•The root means square error predicted by the GANN model was minimal.•HAU-M1 combined with EH shows improved hydrogen production potential.