An enhanced intrusion detection method for AIM of smart grid

As a highly automated power transmission network, the smart grid can monitor each user and grid node and connect different devices to improve the function of conventional power network significantly, but this heterogeneous network also brings greater security risks, attackers can use vulnerabilities existing in smart grids. Intrusion Detection System (IDS) constitutes an important means to protect critical information from being leaked. in a smart grid environment. In this paper, we proposed an AMI intrusion detection model for smart grid, which is widely distributed in the three-layer architecture of the grid system through particle swarm algorithm combined with random forest method. To improve the model’s accuracy, this paper adopts the dynamic weight formula and various adaptive mutation methods to optimize the iterative process of the algorithm. Besides, we use parallel strategy to make up for the lack of precision in the mutation of the algorithm. The AM-PPSO algorithm proposed in this paper performs well in the CEC2017 benchmark function test, effectively ensuring the improvement of the RF classifier. Finally, we use NPL-KDD, UNSW-UB15, and X-IIoTID standard intrusion detection datasets to simulate, results show that our model achieves 97–99% classification of the three datasets.