In this study, blends of 5 % and 10 % ethanol with waste cooking oil biodiesel are mixed with Multi-Walled Carbon Nanotubes (MWCNT) at a concentration of 30 ppm. MWCNTs, known for their high surface area and unique properties, are added to potentially enhance combustion efficiency and reduce emissions. These blends are then tested in a diesel engine to evaluate their performance, emission and combustion characteristics using Design of experiment technique. After conducting various experiments and analyses, the blend of 20 % biodiesel, 10 % ethanol, and 30 ppm MWCNT was identified as the most optimal due to its favorable engine characteristics. Brake Thermal Efficiency (BTE) was increased from 3.1 % to 3.4 % with the addition of MWCNTs, indicating enhanced fuel efficiency. Moreover, average Fuel Consumption is decreased from 2.2 % to 2.5 %, suggesting improved fuel utilization. Using MWCNT 30 ppm in B20 ethanol blends (MWCNT 30 ppm B20+E10) resulted in 35 % reduction in nitrogen oxide (NOx) emissions, 37 % reduction in CO emissions and 39 % reduction in HC emissions. Hence MWCNTs demonstrated effectiveness in mitigating harmful exhaust emissions The optimized values for all parameters fall within acceptable ranges, indicated successful optimization using Response surface methodology. Additionally, statistical analysis reveals that the machine learning- XGBoost model outperformed all other advanced machine learning models across all tested metrics, including MSE, MAE, and R-square.

中文翻译：

---

使用废弃食用油-乙醇生物柴油与 MWCNT 纳米粒子的混合物对柴油发动机的性能、排放和燃烧特性进行实验研究


在这项研究中，将 5% 和 10% 乙醇与废弃食用油生物柴油的混合物与浓度为 30 ppm 的多壁碳纳米管 (MWCNT) 混合。多壁碳纳米管以其高表面积和独特的性能而闻名，添加它可以潜在地提高燃烧效率并减少排放。然后使用实验设计技术在柴油发动机中测试这些混合物，以评估其性能、排放和燃烧特性。经过各种实验和分析，20% 生物柴油、10% 乙醇和 30 ppm MWCNT 的混合物因其良好的发动机特性而被认为是最佳混合物。添加多壁碳纳米管后，制动热效率 (BTE) 从 3.1% 增加到 3.4%，这表明燃油效率得到提高。此外，平均燃料消耗从 2.2% 降低至 2.5%，表明燃料利用率得到提高。在 B20 乙醇混合物（MWCNT 30 ppm B20+E10）中使用 MWCNT 30 ppm 可使氮氧化物 (NOx) 排放量减少 35%，CO 排放量减少 37%，HC 排放量减少 39%。因此，多壁碳纳米管在减少有害废气排放方面表现出有效性。所有参数的优化值均落在可接受的范围内，表明使用响应面方法进行了成功的优化。此外，统计分析表明，机器学习 XGBoost 模型在所有测试指标（包括 MSE、MAE 和 R 方）上均优于所有其他高级机器学习模型。