Scholarlink

Energy management in healthcare facilities is critical for operational efficiency and environmental sustainability. This study investigates the use of Long Short-Term Memory (LSTM) neural networks to predict future energy consumption using a real-world dataset of hourly electricity usage from a hospital in Phoenix, USA. The dataset comprises 8,760 instances spanning an entire year, enabling detailed time-series analysis. Exploratory data analysis reveals pronounced daily and weekly seasonal patterns in electricity consumption, reflecting the typical operational dynamics of healthcare facilities. The LSTM model is leveraged for its ability to capture long-term dependencies and complex temporal patterns in sequential data. By training the model on this dataset, we achieve accurate predictions of future energy consumption, providing valuable insights for energy planning and resource allocation. The results demonstrate the efficacy of LSTM in forecasting energy usage trends, outperforming traditional statistical models in capturing non-linear and seasonal behaviours. This research highlights the potential of deep learning-based approaches to optimize energy efficiency, reduce operational costs, and mitigate environmental impacts in critical infrastructure like hospitals. The findings serve as a foundation for implementing intelligent energy management systems in similar settings, promoting sustainability while ensuring uninterrupted healthcare services. The evaluation of the proposed system demonstrates its effectiveness in accurately predicting hospital energy consumption using LSTM networks. The model’s performance was assessed through metrics such as Mean Squared Error (MSE) and Root Mean Squared Error (RMSE), achieving high accuracy in capturing time-dependent patterns and fluctuations in energy usage. By leveraging real-world hospital data, the system proved reliable in forecasting energy demands, enabling better resource planning and operational efficiency. Additionally, its ability to deliver real-time predictions and optimize energy usage highlights its potential to reduce costs and support sustainability goals in healthcare facilities.

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

Energy Consumption Time Series Forecasting Using LSTM Model

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • I. Hemalatha,
  • Kari Jyothi Chandrika,
  • Pericharla Durga Sahithi,
  • Rasabathula Bhargavi,
  • Anupoju Annapurneswari,
  • Manda Varsheta Swaroopa

摘要

Energy management in healthcare facilities is critical for operational efficiency and environmental sustainability. This study investigates the use of Long Short-Term Memory (LSTM) neural networks to predict future energy consumption using a real-world dataset of hourly electricity usage from a hospital in Phoenix, USA. The dataset comprises 8,760 instances spanning an entire year, enabling detailed time-series analysis. Exploratory data analysis reveals pronounced daily and weekly seasonal patterns in electricity consumption, reflecting the typical operational dynamics of healthcare facilities. The LSTM model is leveraged for its ability to capture long-term dependencies and complex temporal patterns in sequential data. By training the model on this dataset, we achieve accurate predictions of future energy consumption, providing valuable insights for energy planning and resource allocation. The results demonstrate the efficacy of LSTM in forecasting energy usage trends, outperforming traditional statistical models in capturing non-linear and seasonal behaviours. This research highlights the potential of deep learning-based approaches to optimize energy efficiency, reduce operational costs, and mitigate environmental impacts in critical infrastructure like hospitals. The findings serve as a foundation for implementing intelligent energy management systems in similar settings, promoting sustainability while ensuring uninterrupted healthcare services. The evaluation of the proposed system demonstrates its effectiveness in accurately predicting hospital energy consumption using LSTM networks. The model’s performance was assessed through metrics such as Mean Squared Error (MSE) and Root Mean Squared Error (RMSE), achieving high accuracy in capturing time-dependent patterns and fluctuations in energy usage. By leveraging real-world hospital data, the system proved reliable in forecasting energy demands, enabling better resource planning and operational efficiency. Additionally, its ability to deliver real-time predictions and optimize energy usage highlights its potential to reduce costs and support sustainability goals in healthcare facilities.