Scholarlink

<p>Northeast India, with its abundant rainfall, unreliable electricity, and extensive crop cultivation, necessitates reliable and renewable drying technologies to minimize post-harvest losses. This study focuses on the development and performance evaluation of a biomass-fueled natural convection dryer integrated with thermal energy storage materials—paraffin wax and pebbles—for drying ginger. Experimental and computational analyses reveal consistent airflow and temperature distribution, with the combination of both storage materials offering maximum heat retention and paraffin wax showing the most efficient moisture removal. The system achieves a drying efficiency of approximately 40%. To enhance predictive capabilities, a machine learning approach was employed using experimental data generated under four operational scenarios. Four algorithms—decision tree, K-nearest neighbors (KNNs), random forest and gradient boosting regressor (GBR)—were trained and compared. In single-output regression, GBR provided the most accurate enthalpy prediction (R<sup>2</sup> = 0.9820), while KNN outperformed others for energy efficiency (R<sup>2</sup> = 0.9423), exergy efficiency (R<sup>2</sup> = 0.9714), and exergy loss (R<sup>2</sup> = 0.9760). In multi-output regression, GBR yielded the best performance with an R<sup>2</sup> of 0.9657. The integration of experimental validation, simulation, and machine learning modeling demonstrates a comprehensive and robust framework for improving the efficiency and applicability of biomass-based drying systems in rural and hilly agricultural regions.</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

Computational and experimental analysis of a biomass-fueled natural convection dryer and prediction of thermodynamic parameters using machine learning algorithms

  • 期刊论文

  • 发表日期：

    2025年10月9日

查看全文
  • Arnab Deb,
  • Rahul Pache,
  • Manjula Das Ghatak,
  • Ranjan Kumar,
  • Sandip Kumar Mandal

摘要

Northeast India, with its abundant rainfall, unreliable electricity, and extensive crop cultivation, necessitates reliable and renewable drying technologies to minimize post-harvest losses. This study focuses on the development and performance evaluation of a biomass-fueled natural convection dryer integrated with thermal energy storage materials—paraffin wax and pebbles—for drying ginger. Experimental and computational analyses reveal consistent airflow and temperature distribution, with the combination of both storage materials offering maximum heat retention and paraffin wax showing the most efficient moisture removal. The system achieves a drying efficiency of approximately 40%. To enhance predictive capabilities, a machine learning approach was employed using experimental data generated under four operational scenarios. Four algorithms—decision tree, K-nearest neighbors (KNNs), random forest and gradient boosting regressor (GBR)—were trained and compared. In single-output regression, GBR provided the most accurate enthalpy prediction (R2 = 0.9820), while KNN outperformed others for energy efficiency (R2 = 0.9423), exergy efficiency (R2 = 0.9714), and exergy loss (R2 = 0.9760). In multi-output regression, GBR yielded the best performance with an R2 of 0.9657. The integration of experimental validation, simulation, and machine learning modeling demonstrates a comprehensive and robust framework for improving the efficiency and applicability of biomass-based drying systems in rural and hilly agricultural regions.