ANN-ANFIS model for optimising methylic composite biodiesel from neem and castor oil and predicting emissions of the biodiesel blend

Researchers and stakeholders have shown interest in heterogeneous composite biodiesel (HCB) due to its enhanced fuel properties and environmental friendliness (EF). The lack of high viscosity datasets for parent hybrid oils has hindered their commercialisation. Reliable models are lacking to optimis...

Full description

Bibliographic Details
Main Authors: Chao, Zhe Zhu, Samuel, Olusegun David, Taheri-Garavand, Amin, Elboughdiri, Noureddine, Paramasivam, Prabhu, Hussain, Fayaz, Enweremadu, Christopher C., Ayanie, Abinet Gosaye
Format: Article
Language:English
Published: Nature Research 2025
Online Access:http://psasir.upm.edu.my/id/eprint/120193/
http://psasir.upm.edu.my/id/eprint/120193/1/120192.pdf
_version_ 1848868135055130624
author Chao, Zhe Zhu
Samuel, Olusegun David
Taheri-Garavand, Amin
Elboughdiri, Noureddine
Paramasivam, Prabhu
Hussain, Fayaz
Enweremadu, Christopher C.
Ayanie, Abinet Gosaye
author_facet Chao, Zhe Zhu
Samuel, Olusegun David
Taheri-Garavand, Amin
Elboughdiri, Noureddine
Paramasivam, Prabhu
Hussain, Fayaz
Enweremadu, Christopher C.
Ayanie, Abinet Gosaye
author_sort Chao, Zhe Zhu
building UPM Institutional Repository
collection Online Access
description Researchers and stakeholders have shown interest in heterogeneous composite biodiesel (HCB) due to its enhanced fuel properties and environmental friendliness (EF). The lack of high viscosity datasets for parent hybrid oils has hindered their commercialisation. Reliable models are lacking to optimise the transesterification parameters for developing HCB, and the scarcity of predictive models has affected climate researchers and environmental experts. In this study, basic fuel properties were analysed, and models were developed models for the yield of HCB and kinematic viscosity (KV) for composite biodiesel/neem castor seed oil methyl ester (NCSOME) using Artificial Neural Network (ANN) and Adaptive Neuro Fuzzy Inference System (ANFIS). Statistical indices such as computed coefficient of determination (R2), root-mean-square-error (RMSE), standard error of prediction (SEP), mean average error (MAE), and average absolute deviation (AAD) were used to evaluate the effectiveness of the techniques. Emission models for NCSOME-diesel blends were also established. The study investigated the impact of optimised fuel types/NCSOME-diesel (10–30 vol%), ZnO nanoparticle dosage (400–800 ppm), engine speed (1100–1700 rpm), and engine load (10–30%) on emission characteristics and environmental friendliness indices (EFI) such as carbon monoxide (CO), Oxides of Nitrogen (NOx), and Unburnt Hydrocarbon (UHC) using Response Surface Methodology (RSM). The ANFIS model demonstrated superior performance in terms of R2, RMSE, SEP, MAE, and AAD compared to the ANN model in predicting yield and KV of HCB. The optimal emission levels for CO (49.26 ppm), NOx (0.5171 ppm), and UHC (2.783) were achieved with a fuel type of 23.4%, nanoparticle dosage of 685.432 ppm, engine speed of 1329.2 rpm, and engine load of 10% to ensure cleaner EFI. The hybrid ANFIS and ANN models can effectively predict and model fuel-related characteristics and improve the HCB process, while the RSM model can be a valuable tool for climate and environmental stakeholders in accurate forecasting and promoting a cleaner environment. The valuable datasets can also provide reliable information for strategic planning in the biodiesel and automotive industries.
first_indexed 2025-11-15T14:47:34Z
format Article
id upm-120193
institution Universiti Putra Malaysia
institution_category Local University
language English
last_indexed 2025-11-15T14:47:34Z
publishDate 2025
publisher Nature Research
recordtype eprints
repository_type Digital Repository
spelling upm-1201932025-09-25T00:15:55Z http://psasir.upm.edu.my/id/eprint/120193/ ANN-ANFIS model for optimising methylic composite biodiesel from neem and castor oil and predicting emissions of the biodiesel blend Chao, Zhe Zhu Samuel, Olusegun David Taheri-Garavand, Amin Elboughdiri, Noureddine Paramasivam, Prabhu Hussain, Fayaz Enweremadu, Christopher C. Ayanie, Abinet Gosaye Researchers and stakeholders have shown interest in heterogeneous composite biodiesel (HCB) due to its enhanced fuel properties and environmental friendliness (EF). The lack of high viscosity datasets for parent hybrid oils has hindered their commercialisation. Reliable models are lacking to optimise the transesterification parameters for developing HCB, and the scarcity of predictive models has affected climate researchers and environmental experts. In this study, basic fuel properties were analysed, and models were developed models for the yield of HCB and kinematic viscosity (KV) for composite biodiesel/neem castor seed oil methyl ester (NCSOME) using Artificial Neural Network (ANN) and Adaptive Neuro Fuzzy Inference System (ANFIS). Statistical indices such as computed coefficient of determination (R2), root-mean-square-error (RMSE), standard error of prediction (SEP), mean average error (MAE), and average absolute deviation (AAD) were used to evaluate the effectiveness of the techniques. Emission models for NCSOME-diesel blends were also established. The study investigated the impact of optimised fuel types/NCSOME-diesel (10–30 vol%), ZnO nanoparticle dosage (400–800 ppm), engine speed (1100–1700 rpm), and engine load (10–30%) on emission characteristics and environmental friendliness indices (EFI) such as carbon monoxide (CO), Oxides of Nitrogen (NOx), and Unburnt Hydrocarbon (UHC) using Response Surface Methodology (RSM). The ANFIS model demonstrated superior performance in terms of R2, RMSE, SEP, MAE, and AAD compared to the ANN model in predicting yield and KV of HCB. The optimal emission levels for CO (49.26 ppm), NOx (0.5171 ppm), and UHC (2.783) were achieved with a fuel type of 23.4%, nanoparticle dosage of 685.432 ppm, engine speed of 1329.2 rpm, and engine load of 10% to ensure cleaner EFI. The hybrid ANFIS and ANN models can effectively predict and model fuel-related characteristics and improve the HCB process, while the RSM model can be a valuable tool for climate and environmental stakeholders in accurate forecasting and promoting a cleaner environment. The valuable datasets can also provide reliable information for strategic planning in the biodiesel and automotive industries. Nature Research 2025 Article PeerReviewed text en cc_by_4 http://psasir.upm.edu.my/id/eprint/120193/1/120192.pdf Chao, Zhe Zhu and Samuel, Olusegun David and Taheri-Garavand, Amin and Elboughdiri, Noureddine and Paramasivam, Prabhu and Hussain, Fayaz and Enweremadu, Christopher C. and Ayanie, Abinet Gosaye (2025) ANN-ANFIS model for optimising methylic composite biodiesel from neem and castor oil and predicting emissions of the biodiesel blend. Scientific Reports, 15 (1). art. no. 5638. pp. 1-27. ISSN 2045-2322 https://www.nature.com/articles/s41598-025-88901-9?error=cookies_not_supported&code=7b3c9eed-fadb-4ecb-a00c-6c6f7aa80e20 10.1038/s41598-025-88901-9
spellingShingle Chao, Zhe Zhu
Samuel, Olusegun David
Taheri-Garavand, Amin
Elboughdiri, Noureddine
Paramasivam, Prabhu
Hussain, Fayaz
Enweremadu, Christopher C.
Ayanie, Abinet Gosaye
ANN-ANFIS model for optimising methylic composite biodiesel from neem and castor oil and predicting emissions of the biodiesel blend
title ANN-ANFIS model for optimising methylic composite biodiesel from neem and castor oil and predicting emissions of the biodiesel blend
title_full ANN-ANFIS model for optimising methylic composite biodiesel from neem and castor oil and predicting emissions of the biodiesel blend
title_fullStr ANN-ANFIS model for optimising methylic composite biodiesel from neem and castor oil and predicting emissions of the biodiesel blend
title_full_unstemmed ANN-ANFIS model for optimising methylic composite biodiesel from neem and castor oil and predicting emissions of the biodiesel blend
title_short ANN-ANFIS model for optimising methylic composite biodiesel from neem and castor oil and predicting emissions of the biodiesel blend
title_sort ann-anfis model for optimising methylic composite biodiesel from neem and castor oil and predicting emissions of the biodiesel blend
url http://psasir.upm.edu.my/id/eprint/120193/
http://psasir.upm.edu.my/id/eprint/120193/
http://psasir.upm.edu.my/id/eprint/120193/
http://psasir.upm.edu.my/id/eprint/120193/1/120192.pdf