Scholarlink

<p>The inherent inefficiency of conventional solar air collectors (SACs) due to poor heat exchange between the absorber plate (AP) and air necessitates innovative solutions. The present experimental study investigates an enhanced SAC design (Type B) incorporating used aluminum cans filled with sensible heat storage (SHS) attached to the AP. A comparative performance analysis was conducted against a traditional SAC (Type A) under various mass flow rates (0.02–0.08&#xa0;kg s<sup>−1</sup>) and inclination angles (30° and 45°). The integration of SHS not only aims to augment heat transfer but also to extend the effective operational period of the SAC. Results indicate that the modified SAC achieved its optimal performance at a mass flow rate of 0.02&#xa0;kg s<sup>−1</sup> and a 45° tilt angle, leading to significant improvements of 13.6% in outlet temperature and 15.1% in thermal efficiency. To further analyze the system, five machine learning models (XGBoost, SVR, AdaBoost, RF, and KNN) were used to predict outlet temperature and thermal efficiency, with the XGBoost algorithm demonstrating superior predictive capability.</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

Thermodynamic performance analysis of sensible energy storage assisted solar air collector leveraging machine learning models

  • 期刊论文

  • 发表日期：

    2026年2月23日

查看全文
  • Pranjal Prasad Newar,
  • Sujit Roy,
  • Subhankar Saha,
  • Suresh Gogada,
  • Biplab Das,
  • Abhijit Bhowmik

摘要

The inherent inefficiency of conventional solar air collectors (SACs) due to poor heat exchange between the absorber plate (AP) and air necessitates innovative solutions. The present experimental study investigates an enhanced SAC design (Type B) incorporating used aluminum cans filled with sensible heat storage (SHS) attached to the AP. A comparative performance analysis was conducted against a traditional SAC (Type A) under various mass flow rates (0.02–0.08 kg s−1) and inclination angles (30° and 45°). The integration of SHS not only aims to augment heat transfer but also to extend the effective operational period of the SAC. Results indicate that the modified SAC achieved its optimal performance at a mass flow rate of 0.02 kg s−1 and a 45° tilt angle, leading to significant improvements of 13.6% in outlet temperature and 15.1% in thermal efficiency. To further analyze the system, five machine learning models (XGBoost, SVR, AdaBoost, RF, and KNN) were used to predict outlet temperature and thermal efficiency, with the XGBoost algorithm demonstrating superior predictive capability.