| VISION | The goal is to enable real time monitoring of coil performance in wireless charging scenarios to detect external attacks or abnormal interference. Furthermore, by comparing the performance of different algorithm models, the aim is to identify the most stable and reliable detection solution. Using a unified dataset and evaluation metrics, each member designs a model, and the advantages and disadvantages of each model are compared to provide a basis for the wireless charging safety detection baseline and algorithm library. |
|---|---|
| MISSION | This project implements a prototype system of a “single-sample strongly stable classifier.” Starting with the raw signal reading, preprocessing, and data augmentation of the wireless charging coil, it utilizes a one-dimensional convolutional neural network to extract features and perform binary classification on a single time series, automatically determining whether the current charging is normal or if an attack is possible. Then, this model is systematically compared with other members’ models on the same dataset. |
| TEAM | Guangpuzhao Yang, Hanwen Zhu, Hang Liu, Yifan Zhang |
| CLIENT | Dr. Xianglong Feng |
| SITE |