Evaluation of Machine Learning Models for Smart Grid Parameters - Performance Analysis of ARIMA and Bi-LSTM

Aug 4, 2026

Renewable energy integration into smart grids is critical for managing and forecasting energy production in the fourth energy revolution. This study applies machine learning, specifically ARIMA and bidirectional long short-term memory (Bi-LSTM), to predict next-year solar power output. Trained and tested on one year of real-time solar generation data, the models are evaluated using mean absolute error (MAE) and root mean squared error (RMSE). Bi-LSTM achieves higher accuracy than ARIMA and captures complex patterns and long-term dependencies in real-time series data, supporting more efficient, sustainable power systems.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages25

Reading Time1 h 3 min

AccessFree to download

Introduction

Renewable energy and smart grid forecasting

Machine learning approaches for prediction

Methodology

ARIMA model

Bi-LSTM model

Results and Performance Analysis

Error metrics (MAE, RMSE)

Model comparison

Conclusion

sustainability Article Evaluation of Machine Learning Models for Smart Grid Parameters: Performance Analysis of ARIMA and Bi-LSTM Yuanhua Chen 1,2, Muhammad Shoaib Bhutta 2, *, Muhammad Abubakar 3,*, Dingtian Xiao 2, Fahad M. Almasoudi 4, Hamad Naeem 5 and Muhammad Faheem 6 Citation: Chen, Y.; Bhutta, M.S.; Abubakar, M.; Xiao, D.; Almasoudi, F.M.; Naeem, H.; Faheem, M. Evaluation of Machine Learning Models for Smart Grid Parameters: Performance Analysis of ARIMA and Bi-LSTM. Sustainability 2023, 15, 8555 . [https://doi.org/10.3390/su15118555](https://doi.org/10.3390/su15118555) Academic Editor: Luis Hern¡ndez-Callejo Received: 8 April 2023 Revised: 18 May 2023 Accepted: 22 May 2023 Published: 25 May 2023 Copyright: © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license ([https://](https://)[ ](https://)[creativecommons.org/licenses/by/](creativecommons.org/licenses/by/)[ ](creativecommons.org/licenses/by/)[4.0/](4.0/)) . 1 College of Mechanical and Vehicle Engineering, Hunan University, Changsha 410082, China; [chenyuanhua202209@163.com](chenyuanhua202209@163.com) 2 School of Automobile Engineering, Guilin University of Aerospace Technology, Guilin 541004, China; [18487356130@163.com](18487356130@163.com) 3 Key Laboratory of Smart Grid of Ministry of Education, Tianjin University, Tianjin 300072, China 4 Department of Electrical Engineering, Faculty of Engineering, University of Tabuk, Tabuk 47913, Saudi Arabia; [falmasoudi@ut.edu.sa](falmasoudi@ut.edu.sa) 5 Department of Computer Science, King Faisal University, Hofuf 31982, Saudi Arabia; [hamadnaeemh@yahoo.com](hamadnaeemh@yahoo.com) 6 Department of Computing Technology and Innovations, University of Vaasa, 65200 Vaasa, Finland; muhammad.faheem@uwasa.ﬁ * Correspondence: [shoaib@guat.edu.cn](shoaib@guat.edu.cn) (M.S.B.); [mr_abubakar58@hotmail.com](mr_abubakar58@hotmail.com) (M.A.) Abstract: The integration of renewable energy resources into smart grids has become increasingly important to address the challenges of managing and forecasting energy production in the fourth energy revolution. To this end, artiﬁcial intelligence (AI) has emerged as a powerful tool for improving energy production control and management. This study investigates the application of machine learning techniques, speciﬁcally ARIMA (auto-regressive integrated moving average) and Bi-LSTM (bidirectional long short-term memory) models, for predicting solar power production for the next year. Using one year of real-time solar power production data, this study trains and tests these modelson performance measures such as mean absolute error (MAE) and root mean squared error (RMSE) . The results demonstrate that the Bi-LSTM (bidirectional long short-term memory) model outperforms the ARIMA (auto-regressive integrated moving average) model in terms of accuracy and is able to successfully identify intricate patterns and long-term relationships in the real-time-series data. The ﬁndings suggest that machine learning techniques can optimize the integration of renewable energy resources into smart grids, leading to more efﬁcient and sustainable power systems. Keywords: renewable energy; smart grids; energy forecasting; ARIMA; Bi-LSTM model 1. Introduction As the world is moving towards sustainability, there is an increasing need for renewable and green energy resources. Non-renewable resources such as fossil fuels are rapidly disappearing, and their consumption is causing severe environmental hazards. Therefore, renewable energy resources such as solar energy, wind energy, and hydroelectric power are becoming more and more popular [1] . However, the integration of these resources into the smart grid is a challenge that needs to be addressed. Almost all researchers focus on enhancing stability and control in power grid systems through th

FAQs

What machine learning models are compared in the study?

The study compares ARIMA with a bidirectional long short-term memory (Bi-LSTM) model for solar power prediction.

What data and evaluation metrics are used?

Models are trained and tested using one year of real-time solar power production data and evaluated with MAE and RMSE.

Which model performs better and why?

Bi-LSTM outperforms ARIMA in accuracy, because it can identify intricate patterns and long-term relationships in the real-time time series.

Aug 4, 2026

Upload by Olivia Brown