Scholarlink

<p>Internal combustion engines (ICEs) play important roles in transportation and electricity generation, but also contribute significantly to fossil fuel depletion and greenhouse gas emissions. Biogas, a renewable fuel with methane and carbon dioxide, offers a sustainable fuel source for ICEs. However, biogas variable composition and complex, non-linear combustion behavior can lead to unstable operation, inefficient performance, and increased emissions. To tackle these issues, this research presents a novel intelligent optimization and prediction framework, the Multiplayer Battle Game-Inspired Optimizer (MBGIO) and the Multi-Component Attention Graph Convolutional Neural Network (MCAGCNN), known as the MBGIO-MCAGCNN framework. The MBGIO algorithm optimizes key engine parameters (EP) like compression ratio, engine load, injection pressure, and biogas flow rate by simulating multiplayer game dynamics that balance exploration and exploitation, thereby enhancing efficiency and reducing emissions. Meanwhile, the MCAGCNN model uses spatial and temporal interdependence to reliably predict engine vibration, noise, and emission characteristics. This integration allows for simultaneous optimization and prediction of engine performance, resulting in improved stability and environmental compliance. The proposed methodology was implemented in MATLAB and tested with the metrics Efficiency, Mean Squared Error (MSE), and Mean Absolute Error (MAE). Simulation outcomes reveal that the proposed MBGIO-MCAGCNN framework achieved a maximum efficiency of 99.5% with the lowest prediction errors observed through an MSE of 15.65% and an MAE of 14.34% compared to existing methods, including Multi-Input Multi-Output and Artificial Neural Network (MIMO-ANN), Multi-Output Least-Squares Support Vector Regression (MLS-SVR), Autoregressive Integrated Moving Average (ARIMA), Multi-Objective Optimization based on Response Surface Methodology (MOO-RSM), and Artificial Neural Network (ANN). The findings demonstrated that the proposed framework enhanced combustion stability, pollution control, and operational performance in biogas-fuelled ICE.</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

Optimizing biogas-fuelled internal combustion engine performance and emission reduction using a hybrid intelligent modeling approach

  • 期刊论文

  • 发表日期：

    2026年3月12日

查看全文
  • M. Ravi,
  • M. Sasikumar,
  • B. Dinesh Kumar

摘要

Internal combustion engines (ICEs) play important roles in transportation and electricity generation, but also contribute significantly to fossil fuel depletion and greenhouse gas emissions. Biogas, a renewable fuel with methane and carbon dioxide, offers a sustainable fuel source for ICEs. However, biogas variable composition and complex, non-linear combustion behavior can lead to unstable operation, inefficient performance, and increased emissions. To tackle these issues, this research presents a novel intelligent optimization and prediction framework, the Multiplayer Battle Game-Inspired Optimizer (MBGIO) and the Multi-Component Attention Graph Convolutional Neural Network (MCAGCNN), known as the MBGIO-MCAGCNN framework. The MBGIO algorithm optimizes key engine parameters (EP) like compression ratio, engine load, injection pressure, and biogas flow rate by simulating multiplayer game dynamics that balance exploration and exploitation, thereby enhancing efficiency and reducing emissions. Meanwhile, the MCAGCNN model uses spatial and temporal interdependence to reliably predict engine vibration, noise, and emission characteristics. This integration allows for simultaneous optimization and prediction of engine performance, resulting in improved stability and environmental compliance. The proposed methodology was implemented in MATLAB and tested with the metrics Efficiency, Mean Squared Error (MSE), and Mean Absolute Error (MAE). Simulation outcomes reveal that the proposed MBGIO-MCAGCNN framework achieved a maximum efficiency of 99.5% with the lowest prediction errors observed through an MSE of 15.65% and an MAE of 14.34% compared to existing methods, including Multi-Input Multi-Output and Artificial Neural Network (MIMO-ANN), Multi-Output Least-Squares Support Vector Regression (MLS-SVR), Autoregressive Integrated Moving Average (ARIMA), Multi-Objective Optimization based on Response Surface Methodology (MOO-RSM), and Artificial Neural Network (ANN). The findings demonstrated that the proposed framework enhanced combustion stability, pollution control, and operational performance in biogas-fuelled ICE.