Experimental Evaluation of Regression Prediction Analysis After Testing Engine Performance Characteristics

Using ethanol in gasoline is considered one of the most significant goals in the 2030 agenda, which has been set a 15-year plan in order to achieve it since 2015. Appropriately, this project was planned for predicting the value of the most important engine parameters such as the equivalence air-fuel ratio (φ), fuel consumption (ṁf), and brake thermal efficiency nb. th, and brake-specific fuel consumption (BSFC) by regression models. According to the protocol of this project, first, the determined percentages of ethanol were added (0, 20, 40, 60, and 80%) to gasoline at different engine speeds (850, 1000, 2000, 3000, and 4000 rpm and the New European Driving Cycle test). After testing, calculating, mathematical programming, and fitting the regression models for the two SI-engine (TU5 and EF7) with different properties of engine design,12 regression equations have been determined for each of the ‘ (positive linear model), (ṁf) (negative linear model), nb.th (negative second-order polynomial model), and BSFC (positive second-order polynomial model), respectively. Clearly, these 48 regression equations with different line slopes will be able to predict the exact value of the ‘, (ṁf), nb.th, and BSFC for each concentration of ethanol at different engine speeds in order to help automotive industries for trend predicting them in other similar engines.