Authors:

Seyedarshia Shamszadeh

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

With the increasing demand for owning electric vehicles(EVs), predicting efficient charging solutions are crucial for long-distance travel, especially for autonomous driving. This study focused on estimating Power and Efficiency in a Dynamic Wireless Power Transfer (DWPT) system for electric vehicles using a machine learning approach. This study recognized variations in lateral offset and relative yaw angle could cause significant changes in mutual inductance, directly affecting power transfer efficiency. Thus, a series of parametric simulations were performed in ANSYS Maxwell to generate a dataset of mutual inductance based on various lateral offsets and relative yaw angles. The results were then imported in PLECS to simulate the power transfer. Then an Artificial Neural Network (ANN) model was trained to predict power and efficiency based on alignment variables. The machine learning model enables fast and accurate estimation of Power and Efficiency without the need for repeating simulations which is suitable for prediction of load profile for large scale DWPT. Finally, a real-world simulation between Salt Lake City and Logan was provided to test the accuracy of the machine learning model. Keywords: Load Profile Prediction; Machine Learning; Dynamic Wireless Power Transfer (DWPT); Power Transfer Efficiency; Autonomous Driving

Source:

Purdue University / 2025

Topics:

No topics listed

Co-authors:

Seyedarshia Shamszadeh