Scholarlink

<p>Smart cities (SCs) are complex urban environments that utilize digital technologies and data-driven approaches to improve energy efficiency, environmental sustainability, and infrastructure resilience. The main challenge in smart energy management is the unpredictable and intermittent nature of Renewable Energy Sources (RES), which is responsible for undermining stable energy availability and making grid stability more difficult to attain. Traditional prediction models tend to perform poorly with high errors in forecasting, lesser flexibility toward changing environment conditions, and weak optimization toward real-time integration of energy. To address these difficulties, the research proposed a hybrid deep learning (DL) model, Ensemble Red Fox Optimization with Energy Powered Artificial Neural Network (ERFO-EPANN), intended to precisely forecast solar irradiance and wind speed. A practical dataset was employed, which had solar radiation and meteorological parameters. The information is pre-processed by including z-score normalization, missing value processing, and categorical encoding. The ERFO-EPANN combines bio-inspired optimization with neural learning to enhance forecasting accuracy. It ATtained better results with RMSE and MAE values of 1.650 and 1.100, which surpassed the performance of previous models. This method supports better energy forecasting, maximizes the integration of RES with Energy Storage Systems (ESS), and minimizes fossil fuel dependence. By improving energy reliability and sustainable urban infrastructure, the ERFO-EPANN model positively impacts SCs resilience and environmental sustainability.</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

Empowering smart cities: deep learning for seamless integration of energy storage systems and renewable sources

  • 期刊论文

  • 发表日期：

    2025年10月15日

查看全文
  • Beemkumar Nagappan,
  • Jaymeel Shah,
  • Manali Gupta,
  • Anupam Kumari,
  • B. P. Singh,
  • Satish Upadhyay

摘要

Smart cities (SCs) are complex urban environments that utilize digital technologies and data-driven approaches to improve energy efficiency, environmental sustainability, and infrastructure resilience. The main challenge in smart energy management is the unpredictable and intermittent nature of Renewable Energy Sources (RES), which is responsible for undermining stable energy availability and making grid stability more difficult to attain. Traditional prediction models tend to perform poorly with high errors in forecasting, lesser flexibility toward changing environment conditions, and weak optimization toward real-time integration of energy. To address these difficulties, the research proposed a hybrid deep learning (DL) model, Ensemble Red Fox Optimization with Energy Powered Artificial Neural Network (ERFO-EPANN), intended to precisely forecast solar irradiance and wind speed. A practical dataset was employed, which had solar radiation and meteorological parameters. The information is pre-processed by including z-score normalization, missing value processing, and categorical encoding. The ERFO-EPANN combines bio-inspired optimization with neural learning to enhance forecasting accuracy. It ATtained better results with RMSE and MAE values of 1.650 and 1.100, which surpassed the performance of previous models. This method supports better energy forecasting, maximizes the integration of RES with Energy Storage Systems (ESS), and minimizes fossil fuel dependence. By improving energy reliability and sustainable urban infrastructure, the ERFO-EPANN model positively impacts SCs resilience and environmental sustainability.