Intrusion Detection Approach of Power Information Network Based on DNSAE and IQPSO-SVM

Abstract

As the electric power industry gradually crosses into the new digital comprehensive interconnection era, the power IoT plays an increasingly important role in it. At the same time of the rapid development of power IoT, the power information network responsible for carrying data flow and information flow is at risk of intrusion, and the anomaly detection of network traffic will become an important means to solve this problem. The article proposes a network intrusion detection model based on deep nonsymmetric sparse autoencoder (DNSAE) and improved quantum particle swarm-support vector machine (IQPSO-SVM) to achieve faster and more efficient identification while ensuring accuracy. Firstly, DNSAE is used to extract features from network traffic data, and then these abstracted feature data are used as input for IQPSO-SVM training. The experimental results show that this model has higher detection efficiency than DBN and S-NDAE.