Published Paper Details:

The automobile business has seen tremendous dis- ruption since the introduction of electric vehicles (EVs), which provide a sustainable substitute for traditional internal com- bustion engine automobiles. However, because EVs are more closely linked to digital technologies, they are more vulnerable to hacking. These attacks compromise the reliability, efficiency, and safety of EVs, putting EV owners and their vehicles in grave peril. This article looks at how machine learning techniques can be used to identify and reduce cyberattacks on electric vehicles. the fusion of unsupervised techniques like autoencoders and isolation forests with supervised machine learning algorithms like Random Forest and Support Vector Machine (SVM). A significant amount of data from different EV components, control units, and communica- tion networks will be gathered and analyzed. The multi-layer detection framework improves the accuracy and dependability of cyberattack detection by utilizing the advantages of each technique. While unsupervised algorithms use anomaly detection to identify new or emerging threats, supervised algorithms are trained on labeled datasets to classify current types of cyber- attacks. The suggested model's performance is assessed using a real dataset. This study emphasizes how important machine learning is to protecting the future generation of electric vehicles from new threats and guaranteeing their safe and reliable operation.