Peak load Forecasting using Machine Learning Algorithms
- Resource Type
- Conference
- Authors
- Jain, Akanksha; Gupta, S.C.
- Source
- 2023 IEEE Renewable Energy and Sustainable E-Mobility Conference (RESEM) Renewable Energy and Sustainable E-Mobility Conference (RESEM), 2023 IEEE. :1-4 May, 2023
- Subject
- Engineering Profession
Power, Energy and Industry Applications
Robotics and Control Systems
Transportation
Renewable energy sources
Machine learning algorithms
Load forecasting
Convolution
Machine learning
Predictive models
Software
Power load
Machine Learning
- Language
Machine learning algorithms play varied roles in the forecasting of loads in the electricity sector. The forecasting of electric load depends on multiple nonlinear data, which in turn is responsible for formulating the various energy policies in the generation and distribution sector. Machine learning algorithms and the artificial neural network have performed a noteworthy part in forecasting, and various single and multiple predictive models have been designed so far. This paper proposed a cascaded machine learning algorithm for the prediction of actual load, which is an extension of machine learning and made a comparison with the existing machine learning algorithm and Convolution Neural Networks. The results reveal that Cascaded Machine Learning outperforms load forecasting in comparison to the existing machine learning algorithm and Convolution Neural Networks. This paper analyses short-term load forecasting with some electricity load utilization data gathered from Panama’s power system from 2004 to 2019. The accuracy of forecasting is estimated in terms of Peak Load, RMSE and MAE. MATLAB 2017 software is used for the experimental validation.