Comparative study of machine learning methods integrated with genetic algorithm and particle swarm optimization for bio-char yield prediction

Zeeshan Ul Haq; Hafeez Ullah; Muhammad Nouman Aslam Khan; Salman Raza Naqvi; Abdul Ahad; Nor Aishah Saidina Amin

doi:10.1016/j.biortech.2022.128008

Comparative study of machine learning methods integrated with genetic algorithm and particle swarm optimization for bio-char yield prediction

Bioresour Technol. 2022 Nov:363:128008. doi: 10.1016/j.biortech.2022.128008. Epub 2022 Sep 22.

Authors

Zeeshan Ul Haq¹, Hafeez Ullah¹, Muhammad Nouman Aslam Khan², Salman Raza Naqvi¹, Abdul Ahad¹, Nor Aishah Saidina Amin³

Affiliations

¹ Laboratory of Alternative Fuel and Sustainability, School of Chemical and Materials Engineering, National University of Sciences and Technology, Islamabad 44000, Pakistan.
² Laboratory of Alternative Fuel and Sustainability, School of Chemical and Materials Engineering, National University of Sciences and Technology, Islamabad 44000, Pakistan. Electronic address: mnouman@scme.nust.edu.pk.
³ Chemical Reaction Engineering Group (CREG), School of Chemical & Energy Engineering, Faculty of Engineering, Universiti Teknologi Malaysia (UTM), 81310 Johor Bahru, Malaysia.

PMID: 36155813
DOI: 10.1016/j.biortech.2022.128008

Abstract

In this study, Machine learning (ML) models integrated with genetic algorithm (GA) and particle swarm optimization (PSO) have been developed to predict, evaluate, and analyze biochar yield using biomass properties and process operating conditions. Comparative study of different ML algorithms integrated with GA and PSO were performed to improve the ML models architecture and parameters selection. The results proposed that Ensembled Learning Tree (ELT-PSO) model outperformed all other models and is favored for biochar yield prediction (R² = 0.99, RMSE = 2.33). The partial dependence plots (PDPs) analysis shows the potential effects of each influencing parameter impact on the biochar yield and as well as shows that how these factors will interact during the pyrolysis process. A user-friendly software was developed based on the ELT-PSO model to avoid extensive and expensive experimentations without requiring considerable ML understanding. Difference recorded by GUI was less than 2% with experimental yield.

Keywords: Artificial Intelligence; Biomass; Metaheuristic Techniques; Partial Dependence Analysis; Pyrolysis.

MeSH terms

Algorithms
Charcoal*
Machine Learning*
Pyrolysis

Substances

biochar
Charcoal