A Performance Comparison of Machine Learning Algorithms for Load Forecasting in Smart Grid

Abstract

With the rapid growth of the world\'s population, global electricity demand has skyrocketed. As a result, effective energy management mechanisms are required. Because energy consumption trends are rather volatile. To develop the optimization and control mechanism, precise energy demand estimation and short and/or long-term forecasting results with higher accuracy are required. As a result, machine learning (ML) techniques, in conjunction with distributed demand response programs, are being used to accurately predict future energy demand requirements. In this paper, the performance of various state-of-the-art ML algorithms such as logistic regression (LR), support vector machines (SVM), naive Bayes (NB), decision tree classifier (DTC), K-nearest neighbor (KNN), CatBoost and Extra Tree is examined. The primary goal of this paper is to present a comparison of machine learning (ML) algorithms for short-term load forecasting (STLF) in terms of accuracy and forecast error. Based on the implementation and analysis, we discovered that, when compared to other algorithms, the DTC produces comparatively better results.

Introduction

The rapid growth of the global population and urbanization is driving increased energy demand, especially for electricity, which can be generated through various means such as fossil fuels, nuclear, and renewables. However, traditional power systems (PS) face issues like slow response times, limited monitoring, and vulnerability to failures and cyber-attacks. These challenges, combined with climate change and rising energy needs, highlight the necessity for advanced grid technology.

The Smart Grid (SG) is introduced as a next-generation energy infrastructure offering two-way communication, rapid self-healing capabilities, and better resilience against disasters and cyber threats, unlike the traditional one-way communication electrical grid.

Literature Survey outlines prior works on electricity load forecasting and smart grid control using machine learning techniques.

Existing Methods rely on traditional algorithms like Support Vector Machines (SVM), which suffer from high complexity and slow performance.

Proposed Methods suggest multiple machine learning classifiers—including logistic regression, naive Bayes, decision trees, K-nearest neighbors, CatBoost, and Extra Trees—to improve accuracy, reduce time complexity, and simplify usage.

The Methodology involves using supervised machine learning models to predict electricity usage patterns and detect potential threats in real-time, improving the smart grid’s security and efficiency.

Technologies employed include frontend development (HTML, CSS), backend (Python Django), machine learning libraries (Scikit-learn, XGBoost, TensorFlow/Keras), and database management through Django ORM.

The system features user registration, data upload and preprocessing, model training and selection, and results visualization. Security is ensured via Django’s authentication system.

The platform is a web-based application designed for accessibility and ease of use.

Requirements specify software (Python 3.6+, PyCharm, Flask, pandas, NumPy, Scikit-learn) and hardware (Windows OS, 8GB RAM, 500GB+ storage, mid-to-high end processor).

Results include interface pages for home, about, login, and registration functionalities of the smart grid application.

Conclusion

The SG\'s stability is critical for efficient power distribution to the control stations. ML techniques are crucial in indicating the resilience of the SGs. With the emergence of various Ml algorithms, the primary challenge is to identify the most appropriate algorithm to predict the SG\'s stability. To accomplish this, a comprehensive survey of state-of-the-art ML algorithms for predicting the stability of SGs was conducted. A novel EDTC model is presented in this paper to predict the stability of the smart grid.The proposed model was tested using the NYISO smart grid dataset. EDTC\'s performance is compared to that of traditional ML models such as SVM, KNN, CB,ER, LR, and DT. As part of the future work context aware paradigm, dynamic power requirements can be met while also making the SGs more reliable.

References

[1] B. Zhao, L. Zeng, B. Li, Y. Sun, Z. Wang, M. Shahzad, and P. Xi, ‘‘Collaborative control of thermostatically controlled appliances for balancing renewable generation in smart grid,’’ IEEJ Trans. Electr. Electron. Eng., vol. 15, no. 3, pp. 460–468, Mar. 2020. [2] K. Berk, A. Hoffmann, and A. Müller, ‘‘Probabilistic forecasting of industrial electricity load with regime switching behavior,’’ Int. J. Forecasting, vol. 34, no. 2, pp. 147–162, Apr. 2018. [3] J. R. Cancelo, A. Espasa, and R. Grafe, ‘‘Forecasting the electricity load from one day to one week ahead for the Spanish system operator,’’ Int. J. Forecasting, vol. 24, no. 4, pp. 588–602, 2008. [4] M. Djukanovic, S. Ruzic, B. Babic, D. J. Sobajic, and Y. H. Pao, ‘‘A neuralnet based short term load forecasting using moving window procedure,’’ Int. J. Electr. Power Energy Syst., vol. 17, no. 6, pp. 391–397, Dec. 1995. [5] L. J. Soares and M. C. Medeiros, ‘‘Modeling and forecasting short-term electricity load: A comparison of methods with an application to Brazilian data,’’ Int. J. Forecasting, vol. 24, no. 4, pp. 630–644, Oct. 2008. [6] R. Hu, S. Wen, Z. Zeng, and T. Huang, ‘‘A short-term power load forecasting model based on the generalized regression neural network with decreasing step fruit fly optimization algorithm,’’ Neurocomputing, vol. 221, pp. 24–31, Jan. 2017. [7] A. Kavousi-Fard, H. Samet, and F. Marzbani, ‘‘A new hybrid modified firefly algorithm and support vector regression model for accurate short term load forecasting,’’ Expert Syst. Appl., vol. 41, no. 13, pp. 6047–6056, 2014. [8] Z. H. Osman, M. L. Awad, and T. K. Mahmoud, ‘‘Neural network based approach for short-term load forecasting,’’ in Proc. IEEE/PES Power Syst. Conf. Expo., Mar. 2009, pp. 1–8. [9] N. Zeng, H. Zhang, W. Liu, J. Liang, and F. E. Alsaadi, ‘‘A switching delayed PSO optimized extreme learning machine for short-term load forecasting,’’ Neurocomputing, vol. 240, pp. 175–182, May 2017. [10] V. C. Gungor, D. Sahin, T. Kocak, S. Ergut, C. Buccella, C. Cecati, and G. P. Hancke, ‘‘Smart grid technologies: Communication technologies and standards,’’ IEEE Trans. Ind. Informat., vol. 7, no. 4, pp. 529–539, Nov. 2011. [11] S. K. Desai, A. Dua, N. Kumar, A. K. Das, and J. J. P. C. Rodrigues, ‘‘Demand response management using lattice-based cryptography in smart grids,’’ in Proc. IEEE Global Commun. Conf. (GLOBECOM), Dec. 2018, pp. 1–6. [12] M. Alazab and R. Broadhurst, Cyber-Physical Security: Protecting Critical Infrastructure at the State and Local level, 1st ed, M. C. Robert and H. Simon Eds. Cham, Switzerland: Springer, 2017, pp. 251–266. [13] L. Hernandez, C. Baladron, J. M. Aguiar, B. Carro, A. J. Sanchez-Esguevillas, J. Lloret, and J. Massana, ‘‘A survey on electric power demand forecasting: Future trends in smart grids, microgrids and smart buildings,’’ IEEE Commun. Surveys Tuts., vol. 16, no. 3, pp. 1460–1495, 3rd Quart., 2014. [14] A. Rahman, V. Srikumar, and A. D. Smith, ‘‘Predicting electricity consumption for commercial and residential buildings using deep recurrent neural networks,’’ Appl. Energy, vol. 212, pp. 372–385, Feb. 2018. [15] L. Xiao, W. Shao, C. Wang, K. Zhang, and H. Lu, ‘‘Research and application of a hybrid model based on multi-objective optimization for electrical load forecasting,’’ Energy, vol. 180, pp. 213–233, Oct. 2017.

Copyright

Copyright © 2025 Kaladi Bhanu Prakash Reddy, A Jaya Chandra Reddy, A Varun Kumar Reddy, Kalavakunta Lohith Kumar, Mrs. D Bharathi. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.