An experimental based ANN approach in mapping performance-emission characteristics of a diesel engine operating in dual-fuel mode with LPG

• Diesel-LPG dual fuel reduces NOx, Soot and BSFC with the forfeit of increase in HC and CO emissions.
• Highest LPG injection duration produced maximum HC.
• Lowest LPG injection duration yielded minimum Soot.
• Artificial neural network modeling of BSFC equivalent, CO, NOx, Soot and HC.
• R, R2, MSE, RMSE and MAPE metrics used as evaluation benchmarks.

The present study tries to harness the synergistic exploits of diesel-LPG dual fuel platform coupled with artificial neural network for addressing the environmental intimidation due to pollution, rigorous emission legislatives and the future energy insecurity. The dual fuel operation resulted in higher brake thermal efficiency at high load, and injection duration, which recorded a maximum rise of 11% compared to base line operation. At 25%, 50% and 75% load and optimal dual fuel operation with injection duration of 15,000 μs registered a 52%, 29% and 13% higher BSFC compared to base line diesel operation similar to BSEC. Also higher rates of LPG energy share can be observed with highest injection duration of 15,000 μs. A lower emission rate of 45%, 65%, 27% NOx and Soot is also registered in dual fuel platform with injection duration of 15,000 μs with the penalty of higher HC and CO emission. An ANN model was developed to predict BSFC, BTE, NOx, PM, HC and CO based on the experimental results, with load and injection duration as input parameters for the network. The developed ANN model was capable of predicting the performance and emission parameters with commendable accuracy and resulted in relative values of (R2), RMSE and MAPE of 0.99878, 0.020254 & 4.02% for BSFC, 0.99999, 0.61806 and 0.331536% for NOx 0.991299, 0.013617 & 4.31% for CO 0.99999, 1.24 & 0.443% for HC and finally 0.99918, 0.37011 & 1.71% for Soot.