A Multi-transformer Fusion Intrusion Detection Model for Industrial Internet

With the development of Internet technology, the industrial Internet has penetrated into all aspects of human life, bringing great convenience to human beings. However, the challenge of network security is becoming increasingly serious. Focusing on this problem, this paper proposes a multi-transformer fusion intrusion detection model for real industrial Internet environment, and constructs a WCIDS dataset by collecting real network traffic data. Experiments on multi-class datasets prove that the model achieves high detection accuracy, which is not only applicable to industrial Internet, but also common to other network types.