Prediction of the performance and emissions of a spark ignition engine fueled with butanol-gasoline blends based on support vector regression

Butanol is considered as the more promising alternative fuel candidate because of its favorable chemical and physical properties over ethanol and methanol. In this study, the performance and emissions of a port fuel injected spark ignition engine fueled with butanol-gasoline blends (0-60 vol % butanol blended with gasoline referred as G100-B60), including brake thermal efficiency (BTE), brake specific fuel consumption (BSFC) and carbon monoxide (CO), unburned hydrocarbon (UHC), nitrogen oxide (NOx), were investigated under various equivalence ratio. Among the butanol-gasoline blends, B30 performs well in engine performance and emissions due to its CO (2.3%-8.7%), UHC (12.4%-27.5%), and NOx (2.8%-19.6%) emissions compared to those of gasoline. Butanol can be a good alternative fuel to gasoline for its potential to reduce pollutant emissions. It is well known that engine tests are hard, time consuming, and high cost. Therefore, support vector regression (SVR) was used to predict the performance and emissions of the engine, where equivalence ratio and blend ratio were used as the input parameters, and BTE, BSFC, CO, UHC, and NOx were used as the output parameters. It was observed that the correlation coefficients and mean relative error were in the range of 0.9940-0.9998 and 0.1901-10.2570%, respectively. The SVR predictions of BTE, BSFC, CO, UHC, and NOx yielded the root-mean-squared-errors of 0.0511%, 4.6058 g/kW h, 0.9995% vol, 7.7503 ppm vol and 38.5861 ppm, respectively. It could be indicated that the SVR provided an accurate and simple approach to analyze performance and exhaust emissions of spark ignition engine. (c) 2018 American Institute of Chemical Engineers Environ Prog, 38:e13042, 2019