Hybrid optimization of engine performance and emission using RSM-ANN-GA framework to explore valorization potential of waste cooking oil with green synthesized heterogenous ZnO nanocatalyst

Abstract

The escalating interest in utilizing non-competitive feedstocks for biodiesel production suggests waste cooking oil as a cost-effective resource being explored. The present work highlights the bioconversion of waste cooking oil catalyzed by heterogeneous ZnO nano-catalyst derived from Aloe vera. The performance of DI-CI engine was analyzed for the biodiesel blends of B20 & B20GS50 with externally cooled electronically controlled EGR. The green synthesized zinc oxide nano particle acts as a dual performer including transesterification catalyst and fuel additive. The results revealed that nano catalyzed biodiesel potential of waste cooking oil was 92 ± 0.24 %. The kinetic modelling demonstrated first order reaction kinetics with k = 0.015 min⁻¹ and thermodynamic analysis evidenced endothermic nature of transesterification with ΔH = 39.74 kJ/mol. UV–Vis spectra confirm the presence of zinc oxide nanoparticles at 564 cm⁻¹. FTIR analysis exhibited peaks at 2928, 1442, 3428 and 1750 cm⁻¹ indicate alkane (CH₂), C=O, hydroxyl and amino functional moieties respectively. The sharp peak indicates good crystallinity of the nanoparticles in XRD analysis. Machine learning tools such as RSM and ANN demonstrated lesser RMSE and R² values of 0.1555, 0.2276, 20.33 and 83.54 for BTE and NO_x respectively. The higher R² value of RSM such as 0.9891 and 0.9999 corresponding to BTE and NO_x indicates the reliability of the model towards performance optimization. RSM predicted optimized parameters are 100 % load, 81.564 % biodiesel and 20 % EGR resulted with 31.259 % BTE and 169.2 ppm NO_x. The predictive ability of RSM was higher than ANN. Therefore, the study recommended RSM for performance optimization of DI-CI engine.