Posted in | News | Technology

AI Models Predict Energy Use of Educational Buildings

A recent article published in Scientific Reports proposed four different artificial intelligence (AI) models for predicting the energy consumption of educational buildings. The performance of the models including Decision Trees, K-Nearest Neighbors (KNN), Gradient Boosting, and Long Short-Term Memory (LSTM) network was evaluated and compared in the training and testing stages.

AI Models Predict Energy Use of Educational Buildings
Study: Artificial Intelligence Models Predict School Energy Use. Image Credit: Ground Picture/Shutterstock.com

Background

The building and construction sector accounts for over one-third of global energy demand. Specifically, educational facilities contribute to approximately 37% of the carbon dioxide emissions associated with energy use. Therefore, energy planning and prediction solutions are required for such buildings.

Energy use in school buildings is influenced by various factors including location, size, number of occupants, age, and air conditioning extent. Thus, accurate prediction of energy usage is essential for the effective functioning of contemporary electrical grids. Moreover, it can help develop effective demand-side management plans, intelligent control systems, and fault detection and diagnosis methods.

However, accurate energy consumption prediction is difficult due to several unpredictable situations and noisy data. Consequently, the currently used methods frequently produce inaccurate projections. Thus, this paper developed and validated different AI models to estimate the energy consumption of school buildings using real data.

Methods

The model development process involved multiple stages starting with identifying the factors influencing energy consumption based on literature reviews and experts. Real energy consumption data from 352 educational facilities was filtered to establish a dataset devoid of outliers.

Subsequently, the actual energy consumption data and relevant input parameters were separated into distinct subsets for training, validation, and testing. Overall, 11 input variables and one output variable were selected from the refined dataset.

Descriptive statistics were conducted to obtain concise summaries facilitating comprehension and interpretation of the collected data. Additionally, different visualization tools were employed to identify the patterns, trends, and relationships in the data that would otherwise be hidden. Furthermore, a scatter matrix was constructed to investigate relationships and interactions among multiple variables.

Correlation statistics were employed to investigate the relation between the input variables and the target (annual energy consumption of educational buildings). In addition, parallel coordinate plots were developed for a comprehensive visual representation of multivariate data.

In the second phase of this research, appropriate machine-learning algorithms were selected and trained to estimate the yearly energy usage of educational buildings. The accuracy of these algorithms was determined by parameters such as coefficient of determination (COD), mean absolute error (MAE), mean absolute percentage error (MAPE), and root mean square error (RMSE). Finally, each model’s performance was evaluated using a test dataset.

Results and Discussion

The thorough data analysis revealed interesting relationships between the input parameters and yearly energy usage of educational buildings. For instance, the scatter matrix suggested that larger numbers in the ‘Number of Students', 'Number of Staff', and 'Number of Classrooms' categories could enhance energy use.

However, the effect of physical building characteristics was not as pronounced. The factors such as 'Age of Building' and 'Annual Consumption' also did not straightforwardly impact energy usage. Furthermore, a strong direct relation between the 'Number of Staff' and energy consumption was not observed.

Alternatively, 'AC Capacity' exhibited a high correlation with energy consumption. Moreover, the Pearson Coefficient of Correlation was negative for schools in a city, indicating that certain urban factors might have led to more efficient energy use. The developed AI models performed differently based on their unique characteristics. Tree-based models like Decision Trees and Gradient Boosting handled non-linear relationships effectively, while KNN relied on localized data patterns, and LSTM excelled in capturing temporal dynamics.

The Decision Tree model exhibited strong performance on the training data, with a relatively low RMSE of 20,716.25 and MAE of 10,764.99, indicating a good fit. Additionally, its prediction error was about 3.58% of the actual value, which is decent for practical applications. Alternatively, the KNN model exhibited significantly higher errors, probably indicating overfitting as it had a perfect COD of 0.934134. The RMSE for KNN further increased in the testing phase.

Gradient Boosting outperformed other models during the training phase with a remarkably low RMSE and an almost perfect COD. However, the model's generalization ability was not as high in the testing phase. Overall, the LSTM network performed well on both training and testing data.

Conclusion and Future Prospects

Overall, the researchers successfully developed AI models trained on real data for predicting energy consumption in educational buildings. The factors influencing energy use in such buildings were determined from the literature.

The results demonstrate that by optimizing energy consumption through advanced AI tools, educational facilities can encourage students to engage with ideas of sustainability in their everyday surroundings. Notably, this study included schools located in a hot climate. Thus, the results may not apply to other school types and climates. The researchers suggest conducting an extended study including different school structures and climates in the future.

Journal Reference

Tariq, R., Mohammed, A., Alshibani, A., & Ramírez-Montoya, M. S. (2024). Complex artificial intelligence models for energy sustainability in educational buildings. Scientific Reports, 14(1), 15020. DOI: 10.1038/s41598-024-65727-5, https://www.nature.com/articles/s41598-024-65727-5

Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.

Nidhi Dhull

Written by

Nidhi Dhull

Nidhi Dhull is a freelance scientific writer, editor, and reviewer with a PhD in Physics. Nidhi has an extensive research experience in material sciences. Her research has been mainly focused on biosensing applications of thin films. During her Ph.D., she developed a noninvasive immunosensor for cortisol hormone and a paper-based biosensor for E. coli bacteria. Her works have been published in reputed journals of publishers like Elsevier and Taylor & Francis. She has also made a significant contribution to some pending patents.  

Citations

Please use one of the following formats to cite this article in your essay, paper or report:

  • APA

    Dhull, Nidhi. (2024, July 11). AI Models Predict Energy Use of Educational Buildings. AZoBuild. Retrieved on October 04, 2024 from https://www.azobuild.com/news.aspx?newsID=23560.

  • MLA

    Dhull, Nidhi. "AI Models Predict Energy Use of Educational Buildings". AZoBuild. 04 October 2024. <https://www.azobuild.com/news.aspx?newsID=23560>.

  • Chicago

    Dhull, Nidhi. "AI Models Predict Energy Use of Educational Buildings". AZoBuild. https://www.azobuild.com/news.aspx?newsID=23560. (accessed October 04, 2024).

  • Harvard

    Dhull, Nidhi. 2024. AI Models Predict Energy Use of Educational Buildings. AZoBuild, viewed 04 October 2024, https://www.azobuild.com/news.aspx?newsID=23560.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback
Your comment type
Submit

While we only use edited and approved content for Azthena answers, it may on occasions provide incorrect responses. Please confirm any data provided with the related suppliers or authors. We do not provide medical advice, if you search for medical information you must always consult a medical professional before acting on any information provided.

Your questions, but not your email details will be shared with OpenAI and retained for 30 days in accordance with their privacy principles.

Please do not ask questions that use sensitive or confidential information.

Read the full Terms & Conditions.