Power grid virtual data injection attack detection method based on deep belief network

Abstract

The invention discloses a power grid virtual data injection attack detection method based on a deep belief network, and the method comprises the steps: S1, building a deep belief network detection model which comprises a plurality of layers of RBM networks and a single-layer BP neural network, wherein the RBM networks is connected in a full connection mode; S2, acquiring IEEE standard node measurement data, and adding random noise and attack vectors to the acquired data to generate attack data; s3, dividing the normalized attack data into a test data set and a training data set according to aset proportion; s4, training the RBM network layer by layer from top to bottom by using unsupervised learning, and performing model parameter adjustment through BP neural network back propagation errors to obtain a trained deep belief network detection model; and S5, inputting the test data set into the trained deep belief network detection model, and outputting a prediction result. According to the invention, the dependence of the traditional detection method on the detection threshold is overcome, and the detection of various attack modes is realized.

Description

technical field [0001] The invention relates to the field of smart grids, and more specifically, to a detection method for grid virtual data injection attacks based on a deep belief network. Background technique [0002] As a typical cyber-physical system, smart grid integrates communication, computing and control. While this integration greatly improves the management level of the power system, it also brings greater risks to the system. Power system security incidents have occurred frequently in recent years and have caused great damage. [0003] False data injection attack (FDIA) is an emerging method of data integrity attack. On the basis of a certain understanding of the power network topology, the attacker tampers with the sensor, and the energy management system (EMS) receives this information. The purpose of changing the state estimation result is achieved. Under the action of FDAA, attackers can control the smart devices in the system and manipulate key informati...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the defects of strong dependence on the threshold value of the grid attack detection method and a single detected attack mode in the above-mentioned prior art, the present invention provides a detection method for power grid virtual data injection attack based on a deep trust network

Images