Scholarlink

<p>The development of cost-effective and sustainable heterogeneous catalysts from renewable resources plays a vital role on the biodiesel production process. In this study, Borassus flabellifer biomass was collected, processed and calcined to synthesize a novel heterogeneous catalyst for production of biodiesel by using transesterification process. The characterization studies such as FTIR, TGA, DSC, XRD and FE-SEM were conducted to elucidate its functional groups, crystalline phases, surface morphology, surface area and thermal stability. The transesterification process was chosen to prepare canola biodiesel from raw oil with the aid of a novel catalyst by varying process parameters. The Response Surface Methodology (RSM) was adopted to optimize the operating parameters for obtaining maximum biodiesel yield. Furthermore, machine learning based Random Forest (RF) technique was utilized to model for the prediction of biodiesel yield based on the experimental data. The RSM approach demonstrates an optimum condition of methanol to oil ratio (12:1), reaction temperature (65&#xa0;°C) and catalyst concentration (5 wt %) results a maximum biodiesel yield of 96.24%. The RF model was exhibited a strong predictive accuracy by achieving a high coefficient of determination (R<sup>2</sup> = 0.9785) along with low error values (MAE = 0.3505 and RMSE = 0.4473), indicating its reliability and robustness in predicting biodiesel yield. These findings demonstrate the potential of Borassus biomass as a sustainable heterogeneous catalyst for the biodiesel production and emphasize the role of machine learning based optimization and integration to enhance the biodiesel synthesis.</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

Sustainable Borassus Biomass Derived Catalyst for Biodiesel Production: An Integrated Optimization and Prediction Approach Using RSM and Machine Learning

  • 期刊论文

  • 发表日期：

    2025年11月22日

查看全文
  • Anchupogu Praveen,
  • M. Vimal Teja,
  • Madhavi Katamaneni,
  • Vadlamudi Tara Chand,
  • P. Chengareddy,
  • P. Umamaheswarrao

摘要

The development of cost-effective and sustainable heterogeneous catalysts from renewable resources plays a vital role on the biodiesel production process. In this study, Borassus flabellifer biomass was collected, processed and calcined to synthesize a novel heterogeneous catalyst for production of biodiesel by using transesterification process. The characterization studies such as FTIR, TGA, DSC, XRD and FE-SEM were conducted to elucidate its functional groups, crystalline phases, surface morphology, surface area and thermal stability. The transesterification process was chosen to prepare canola biodiesel from raw oil with the aid of a novel catalyst by varying process parameters. The Response Surface Methodology (RSM) was adopted to optimize the operating parameters for obtaining maximum biodiesel yield. Furthermore, machine learning based Random Forest (RF) technique was utilized to model for the prediction of biodiesel yield based on the experimental data. The RSM approach demonstrates an optimum condition of methanol to oil ratio (12:1), reaction temperature (65 °C) and catalyst concentration (5 wt %) results a maximum biodiesel yield of 96.24%. The RF model was exhibited a strong predictive accuracy by achieving a high coefficient of determination (R2 = 0.9785) along with low error values (MAE = 0.3505 and RMSE = 0.4473), indicating its reliability and robustness in predicting biodiesel yield. These findings demonstrate the potential of Borassus biomass as a sustainable heterogeneous catalyst for the biodiesel production and emphasize the role of machine learning based optimization and integration to enhance the biodiesel synthesis.