Graduate Theses and Dissertations (2019 - present)
Date of Award
8-2025
Document Type
Thesis
Degree Name
M.S.
Department
Electrical and Computer Engineering
Committee Chair
Daniela Wolter Ferreira Touma
Abstract
This paper proposes the prediction of vehicle Driving Energy Consumption (DEC) of based on the user’s driving style. This research focused on the Mobile, AL area. The dataset contains time-series data from a vehicle for six months. All essential data, such as speed, latitude, longitude, and elevation, were collected through the OBD-II port of the vehicles. Before data collection, a survey was conducted between May 2024 and April 2025. 215 participants participated in the survey and shared their viewpoints regarding Electric Vehicles (EVs), Dynamic Pricing (DP) , and driving behaviors. The qualitative survey data showed that the adoption of EVs is not directly related to age, income, or educational level. In addition, a geo-spatial analysis of home and work addresses showed that most of the users leave home early in the morning and return home in the evening. After collecting qualitative and quantitative data from the vehicle, the data was used to calculate the driving energy consumption (DEC). The DEC time-series data was used to train and test machine learning (ML) models to predict the DEC. Then, a logistic regression model is employed to predict the status of the vehicle. Integrating logistic regression with DEC prediction, the home-to-home driving energy (H2HDEC) is calculated for the next day. Robustness analysis of the pretrained models was done with the vehicle’s 7th-month data. The results confirm that the proposed models can capture underlying patterns and provide comparable performance for DEC prediction. The focus on DEC makes the algorithm developed in this research more robust and capable of predicting the energy for any vehicle type because any engine losses are not considered in DEC. The effective energy is calculated considering the average losses for EVs in this research, but the motor losses of any vehicle can be substituted for use with this algorithm.
Recommended Citation
Rahman, Md. Mizanur, "Forecasting Vehicle Energy Consumption based on User Driving Patterns" (2025). Graduate Theses and Dissertations (2019 - present). 226.
https://jagworks.southalabama.edu/theses_diss/226
Included in
Electrical and Electronics Commons, Navigation, Guidance, Control, and Dynamics Commons, Operational Research Commons, Other Electrical and Computer Engineering Commons, Other Engineering Commons, Power and Energy Commons