Title: Towards sustainable transportation and integrating ANN, RSM, and exergy analysis for biofuel-diesel blend optimisation of biodiesel ignition enhancer blends

Authors: S. Ajay; S. Sivani Hanisha; G. Madhan; K. Manikandan

Addresses: Department of Mechanical Engineering, Kongu Engineering College, Tamilnadu, 638060, India ' Department of Mechanical Engineering, Kongu Engineering College, Tamilnadu, 638060, India ' Department of Mechanical Engineering, Kongu Engineering College, Tamilnadu, 638060, India ' Department of Mechanical Engineering, Kongu Engineering College, Tamilnadu, 638060, India

Abstract: To accomplish long-term optimization of biodiesel ignition enhancer blends, this study employs a multi-pronged strategy using artificial neural networks (ANN), response surface methodology (RSM), and exergy analysis. This method methodically investigates parameter space, revealing the best blend ratios that provide the desired efficiency and emissions without compromising on sustainability standards. Approximately 94.3% of the variance in the ignition delay times can be accounted for by the ANN model, and the R-squared value of 0.954 indicates a significant connection between the projected and actual values. The average absolute difference between the model's predicted and observed ignition delay periods is 1.23 milliseconds, demonstrating the model's accuracy. The ANN results are superior to the RSM results in four out of six runs, with a reduced error % compared to the experimental findings. ANNs have a mean squared error (MSE) of 3.17, while RSM is 8.88. Compared to RSM, the ANN provides larger MSEs.

Keywords: ANN; artificial neural networks; exergy; response surface models; biodiesel; ignition enhancer.

DOI: 10.1504/IJEP.2024.142551

International Journal of Environment and Pollution, 2024 Vol.74 No.1/2/3/4, pp.50 - 65

Received: 10 Oct 2023
Accepted: 07 Mar 2024
Published online: 08 Nov 2024 *

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