Scholarlink

<p>The rapid fading of fossils and the importance of fighting global warming has led to the world changing towards the eco-friendly and sustainable sources of energy. However, the management of the intermittency associated with the renewable energy sources coupled with the cost-effectiveness and high performance has become a great challenge. This paper is the techno-economic feasibility study of a combination of renewable energy system (HRES) and green hydrogen production which can full-fill the electricity needs in the buildings. The latest machine learning type of load forecasting is utilized where the Long Short-Term Memory (LSTM) shows high predictive power in comparison to the traditional ones. To achieve optimal system size the Non-Sorting Moth Flame Optimization (NSMFO) algorithm is implemented to trade among performance, cost and resilience. The optimized system is a 5&#xa0;kWp solar photovoltaic, 1.5&#xa0;kW vertical-axis wind generator, and 20&#xa0;kW biogas plant that will run without energy storage system and still deliver reliable and sustainable energy. The findings demonstrate the utility of combining the use of machine learning and superior optimization approaches in the development of the next generation of renewable systems of energy and the furthering of green hydrogen in sustainable infrastructure.</p> Graphical Abstract <p></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

A novel LSTM-NSMFO framework for designing hybrid renewable energy systems with green hydrogen integration

  • 期刊论文

  • 发表日期：

    2026年3月17日

查看全文
  • Murugaperumal Krishnamoorthy,
  • Shanmugasundaram Hariharan,
  • R. Kaladevi,
  • Arulselvi Ramasamy,
  • A. Manju

摘要

The rapid fading of fossils and the importance of fighting global warming has led to the world changing towards the eco-friendly and sustainable sources of energy. However, the management of the intermittency associated with the renewable energy sources coupled with the cost-effectiveness and high performance has become a great challenge. This paper is the techno-economic feasibility study of a combination of renewable energy system (HRES) and green hydrogen production which can full-fill the electricity needs in the buildings. The latest machine learning type of load forecasting is utilized where the Long Short-Term Memory (LSTM) shows high predictive power in comparison to the traditional ones. To achieve optimal system size the Non-Sorting Moth Flame Optimization (NSMFO) algorithm is implemented to trade among performance, cost and resilience. The optimized system is a 5 kWp solar photovoltaic, 1.5 kW vertical-axis wind generator, and 20 kW biogas plant that will run without energy storage system and still deliver reliable and sustainable energy. The findings demonstrate the utility of combining the use of machine learning and superior optimization approaches in the development of the next generation of renewable systems of energy and the furthering of green hydrogen in sustainable infrastructure.

Graphical Abstract