Biodiesel fuel properties and engine characteristics can be improved by using hybrid biodiesel and robust optimization tools. This study predicts the performance parameters of diesel engines fueled with castor-palm kernel biodiesel (CPKB) and diesel fuel blend using response surface methodology (RSM) and artificial neural network (ANN). Optimization of the performance parameters was also carried out using multi-criteria decision-making methods (WASPA S and MOORA) for the first time. The RSM and ANN were employed in predicting the performance parameters such as CPKB fuel blends (FB) (0–20 vol.%), engine load (EL) (0–50%), and engine speed (ES) (1000–2000 rpm) on the performance indicators viz. brake torque (BT), brake power (BP), brake specific fuel consumption (BSFC), brake thermal efficiency (BTE). The Box-Behnken design was used for performing the experimental trials. The RSM model predicted the BT of 107.95 Nm, BP of 11.300 kW, BSFC of 0.057 kg/kWh, BTE of 15.147%, at the optimal level of CPKB blends of 20% (B20), engine load of 50%, and an engine speed of 1000 rpm, respectively. Results showed that based on the values of R2and average absolute deviation (AAD) obtained, the predictive capability of both RSM and ANN were within acceptable limits.
The best experimental trial from the WASPAS method is the #20 experimental run and the parameter combination are FB-10%, EL-25, and ES-1500 rpm, whereas for the MOORA method, five such experimental trials were observed viz., #1 run: FB-0%, EL-0, and ES-1000 rpm, #2 run: FB-20%, EL-0 and ES-1000 rpm, #5 run: FB-0%, EL-0, and ES-2000 rpm, #6 run: FB-20%, EL-0, and ES-2000 rpm #11 run: FB-10%, EL-0, and ES-1500 rpm.