Scholarlink

<p>The purpose of this research is to predict the operation characteristics and emission properties of a diesel engine operating on biodiesel blend containing aluminum oxide (Al2O3) nanoparticles using advanced machine learning (ML) techniques. Three types of Supervised Learning Methods are used to estimate engine operating parameters as follows: Brake Thermal Efficiency (BTE), Brake Specific Fuel Consumption (BSFC), Exhaust Gas Temperature (EGT), CO, HC, CO<sub>2</sub>, NO<sub><i>x</i></sub>: K-Nearest Neighbors (KNN), Random Forest Regressor (RFR), Multi-Layered Perceptron (MLP). The experimental data used to develop the predictive models contained five different biodiesel blends (B20, B40, B60, B80 and B100), each containing two different levels of Al<sub>2</sub>O<sub>3</sub> nanoparticles (2×10<sup>−5</sup> and 4×10<sup>−5</sup>) that were tested within a single cylinder CI engine. Eighty percent of the data collected for this experiment was used to develop and train the predictive models. Twenty percent of the data collected was used to test these predictive models. Statistical evaluation methods such as the Coefficient of Determination (<i>R</i><sup>2</sup>), Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) are used to evaluate the accuracy of the predictive models developed in this study. A high <i>R</i><sup>2</sup> (0.9987) indicates a strong relationship between the actual and predicted values. The RFR technique produced the highest accuracy in comparison to the other two models (KNN and MLP) used in this study as well as indicated that both KNN and MLP were capable of modeling the nonlinear relationships between the combustion process of a diesel engine and its parameters. Adding Al<sub>2</sub>O<sub>3</sub> particles to biodiesel resulted in a significant improvement in the overall fuel characteristics of biodiesel as well as significant reductions in harmful exhaust emissions. Combining experimental data using Artificial Intelligence-based modeling represents a viable and sustainable alternative to improve the operation of diesel engines using renewable biodiesel fuels.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Integrating Multilayer Perceptron, K-Nearest Neighbors and Random Forest Regression Models to Predict Diesel Engine Emissions and Performance with Aluminum Oxide Nanoparticle Biodiesel Blends

  • 期刊论文

  • 发表日期：

    2026年2月16日

查看全文
  • Chetan Pawar,
  • Lokesh V.,
  • Keval Chandrakant Nikam,
  • Laxmikant D. Jathar,
  • Sagar D. Shelare

摘要

The purpose of this research is to predict the operation characteristics and emission properties of a diesel engine operating on biodiesel blend containing aluminum oxide (Al2O3) nanoparticles using advanced machine learning (ML) techniques. Three types of Supervised Learning Methods are used to estimate engine operating parameters as follows: Brake Thermal Efficiency (BTE), Brake Specific Fuel Consumption (BSFC), Exhaust Gas Temperature (EGT), CO, HC, CO2, NOx: K-Nearest Neighbors (KNN), Random Forest Regressor (RFR), Multi-Layered Perceptron (MLP). The experimental data used to develop the predictive models contained five different biodiesel blends (B20, B40, B60, B80 and B100), each containing two different levels of Al2O3 nanoparticles (2×10−5 and 4×10−5) that were tested within a single cylinder CI engine. Eighty percent of the data collected for this experiment was used to develop and train the predictive models. Twenty percent of the data collected was used to test these predictive models. Statistical evaluation methods such as the Coefficient of Determination (R2), Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) are used to evaluate the accuracy of the predictive models developed in this study. A high R2 (0.9987) indicates a strong relationship between the actual and predicted values. The RFR technique produced the highest accuracy in comparison to the other two models (KNN and MLP) used in this study as well as indicated that both KNN and MLP were capable of modeling the nonlinear relationships between the combustion process of a diesel engine and its parameters. Adding Al2O3 particles to biodiesel resulted in a significant improvement in the overall fuel characteristics of biodiesel as well as significant reductions in harmful exhaust emissions. Combining experimental data using Artificial Intelligence-based modeling represents a viable and sustainable alternative to improve the operation of diesel engines using renewable biodiesel fuels.