Title: Hydrogen generation from anaerobic co-digestion and statistical evaluation using machine learning algorithms

Authors: Chinmay Deheri; Saroj Kumar Acharya

Addresses: Department of Mechanical Engineering, ITER, SOA Deemed to be University, Odisha, 751030, India ' Department of Mechanical Engineering, ITER, SOA Deemed to be University, Odisha, 751030, India

Abstract: Hydrogen generation from anaerobic co-digestion of food waste (FW) and cow dung (CD) was statistically predicted using machine learning (ML) models. Laboratory scale experiments were performed using CaO₂ and CaCO₃ as additives. Maximum hydrogen generation of 115.28 and 109.47 mL g^-1 TS was obtained using CaO₂ and CaCO₃. Further, the Pearson correlation matrix evaluated the correlation between the operational parameters such as inoculum to substrate (I/S) ratio, pH, and reactor temperature with the output parameter (hydrogen generation). I/S ratio showed the highest correlation of 0.94 with hydrogen generation compared to the other parameters. Moreover, four regression models were created using ML algorithms such as linear regression (LR), decision tree regression (DTR), random forest regression (RFR), and support vector regression (SVR) to predict hydrogen production. Hydrogen generation was accurately predicted by the ML models with an r² score greater than 0.9 and an RMSE value less than 1.

Keywords: machine learning; ML; anaerobic co-digestion; hydrogen; waste biomass.

DOI: 10.1504/IJGW.2023.130986

International Journal of Global Warming, 2023 Vol.30 No.2, pp.152 - 160

Received: 09 Jun 2022
Accepted: 07 Oct 2022
Published online: 17 May 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article