An intrusion detection model for electric vehicle supply equipment based on multi-task learning and probability transfer mechanism

With the rapid development of new energy vehicles and Internet of Things technology, new energy vehicle users have an increasing demand for electric vehicle supply equipment (EVSE). However, a large number of EVSE are currently vulnerable to remote hacking due to the lack of effective security defense mechanisms. This not only leaks sensitive information of users, but also causes large-scale fluctuations in power grid energy. In addition, current research methods for intrusion detection in EVSE suffer from problems such as high false alarm rate, long detection time, and low detection accuracy. Therefore, in order to build a more effective EVSE intrusion detection method, this paper combines the multi-task learning architecture and proposes a deep learning detection model EVSEMTLIDS to conduct an in-depth analysis of the malicious intrusion behavior of EVSE. In order to shorten the intrusion detection time of this model for multiple tasks, EVSEMTLIDS integrates a multi-task learning architecture. At the same time, we propose a probability transfer mechanism to improve the prediction performance between multiple tasks. Experimental results show that compared with the baseline model, the proposed model improves the detection accuracy by up to 27.29% and shortens the detection time by 43.24%.