Machine learning for enhancing prediction of biogas production and building a VFA/ALK soft sensor in full-scale dry anaerobic digestion of kitchen food waste

Jinlin Zou; Fan Lü; Long Chen; Hua Zhang; Pinjing He

doi:10.1016/j.jenvman.2024.123190

Machine learning for enhancing prediction of biogas production and building a VFA/ALK soft sensor in full-scale dry anaerobic digestion of kitchen food waste

J Environ Manage. 2024 Nov 4:371:123190. doi: 10.1016/j.jenvman.2024.123190. Online ahead of print.

Authors

Jinlin Zou¹, Fan Lü², Long Chen³, Hua Zhang², Pinjing He⁴

Affiliations

¹ State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, 1239 Siping Road, Shanghai, 200092, PR China; Institute of Waste Treatment and Reclamation, Tongji University, Shanghai, 200092, PR China; Shanghai Municipal Engineering Design Institute (Group) Co., Ltd, PR China.
² State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, 1239 Siping Road, Shanghai, 200092, PR China; Institute of Waste Treatment and Reclamation, Tongji University, Shanghai, 200092, PR China; Shanghai Institute of Pollution Control and Ecological Security, Shanghai, 200092, PR China.
³ State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, 1239 Siping Road, Shanghai, 200092, PR China; Institute of Waste Treatment and Reclamation, Tongji University, Shanghai, 200092, PR China.
⁴ State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, 1239 Siping Road, Shanghai, 200092, PR China; Institute of Waste Treatment and Reclamation, Tongji University, Shanghai, 200092, PR China; Shanghai Institute of Pollution Control and Ecological Security, Shanghai, 200092, PR China. Electronic address: [email protected].

PMID: 39504672
DOI: 10.1016/j.jenvman.2024.123190

Abstract

Based on operational data collected over 1.5 years from four full-scale dry anaerobic digesters used for kitchen food waste treatment, this study adopted eight typical machine learning algorithms to distinguish the best biogas prediction model and to develop a soft sensor based on the VFA/ALK ratio. Among all the eight tested models, the CatBoost (CB) algorithm demonstrated superior performance in terms of prediction accuracy and model fitting. Specifically, the CB model achieved a biogas production prediction accuracy (R²) ranging from 0.604 to 0.915, and a VFA/ALK R² between 0.618 and 0.768 on the test dataset. Furthermore, the feature importance analysis revealed that biomass amount into the dry anaerobic digester was the primary factor influencing biogas production. Chemical oxygen demand (COD) and free ammonia nitrogen (FAN) were identified as the most significant factors impacting the VFA/ALK indicator during dry anaerobic digestion, collectively contributing to nearly 50% of the influence. Overall, this study verifies the feasibility of using machine learning to predict biogas production in full-scale dry anaerobic digestion and provides a crucial foundation for monitoring the stability of dry anaerobic digesters.

Keywords: Dry anaerobic digestion; Early warning; Kitchen food waste; Machine learning; Real-time monitoring; Soft sensor.