Modeling of emissions characteristics of a diesel engine fueled by Jatropha Diestrol

Abstract

The global reliance on fossil fuels for energy production has become increasingly apparent. However, the numerous drawbacks and diminishing reserves of fossil fuels have compelled the world to explore and utilize alternative and renewable fuel sources. Biofuels have emerged as a prominent contender among renewable energy options. Derived from animal and plant sources, biofuels obtained from non-edible plant resources have gained precedence to avoid compromising human food supplies. Biodiesel offers several advantages, including clean combustion and energy generation comparable to fossil fuels. One such non-edible plant-based biofuel is Jatropha biodiesel. While a blend of biodiesel and diesel can be directly used in diesel engines, the addition of ethanol can enhance the properties of the fuel blend, resulting in an improved alternative fuel. This advanced fuel blend, known as Jatropha Diesterol, has been developed and patented for the first time in Iran. The Jatropha Diesterol fuel blends (consisting of Jatropha biodiesel-diesel-ethanol) comprised 3% ethanol and 10%, 20%, and 30% biodiesel. These fuel blends underwent testing in a single-cylinder air-cooled diesel engine, operating at full load and four engine speeds (1600, 2000, 2400, and 2800 rpm). The emitted pollutants, namely CO, CO₂, HC, O₂, and HC, were analyzed and recorded. Subsequently, the data was modeled using the support vector machine (SVM) method, incorporating genetic algorithm (GA) optimization. Eighty percent of the data was assigned for training purposes, while the remaining 20% was allocated for testing. The modeling results were evaluated using parameters such as R₂, MSE, MAE, and RSME. The outcomes demonstrated that the SVM+GA method accurately predicted the data from this experiment, achieving high accuracy and, in some cases, 100% accuracy. Therefore, this modeling approach can be utilized for future research in this field, obviating the need for costly and time-consuming experiments and evaluations of advanced alternative fuel blend