Authors:

ez Author

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

This research aims to understand the effects of different variables on electricity demand in the United States based on historical data, prioritizing an environmental context and the influence of electric vehicles (EVs). The study analyzes 13 variables, including energy source generation, electricity consumption, temperature, electric vehicles and economic indicators; each with monthly observations spanning across the last 14 years.?Machine learning supervised iterations, specifically, ridge regression models, are applied to the data. The best model identifies the strongest variables which include summer season, cooling degree days and residential electricity; all of which are coherently related to each other, and also reflected in the peaks of electricity demand.? The study confirms that the EV fleet has a relatively smaller impact on electricity consumption, although it remains a potentially influential market. The study also identifies a clear predominance and reliance for generation of electricity on fossil fuels, especially natural gas, highlighting?the gaps in terms of renewable and mixed energies generation.? Climate change impacts, especially connected to extreme heat, underscore the importance of reducing this dependence on fossil fuels. Even though increased renewable energy generation or a sudden rise in electric vehicle adoption seem like non-conservative scenarios, the existing infrastructure may be insufficient to support these changes. Thereby it is important to consider these scenarios and prepare for them as they could be imminent. Future approaches could include analysis using different machine learning models based on a state or regional level, and the application of autoregressive models for time series forecasting. Keywords: Electricity Demand; Machine Learning; Energy Mix; Electric Vehicles † Presenting Undergrad Author; ‡ Contributing Undergrad Author; * Undergrad Acknowledgment

Source:

Purdue University / 2025

Topics:

No topics listed

Co-authors:

ez Author