FDIA high-precision detection method for power system

Abstract

The invention discloses an FDIA high-precision detection method for a power system. The method comprises the following steps: a historical data packet and power monitoring historical data of the power system are obtained; the GRU is trained by using the training sample set until a loss function value of the GRU is lower than a loss function threshold value so as to obtain a trained GRU state prediction model; the power system receives the data packet and detects power monitoring data; the trained GRU state prediction model obtains a network layer feature estimated value and an electric power monitoring data estimated value of a data packet received by the power system at the tn + 1 moment; the power system actually receives the data packet and actually detects the data packet to obtain power monitoring data; and the GRU state prediction model judges whether the power system is subjected to FDIA or not at the tn + 1 moment. According to the method, the GRU is trained by using the network layer characteristics in the historical data packet and the power monitoring historical data, and the accuracy of judging whether the power system is subjected to FDIA or not by the trained GRU state prediction model is obviously improved.

Description

technical field [0001] The invention relates to an electric power system FDIA high-precision detection method, which belongs to the field of electric power system monitoring methods. Background technique [0002] As the traditional power system becomes increasingly intelligent, its risk of being attacked by information is gradually increasing, among which FDAA (False Data Injection Attack) is a common type of information attack on the power system. For the detection of FDAA, in the prior art, the state value of the power grid is analyzed based on the trained GRU prediction model to determine whether the power system is subject to FDAA. However, for some experienced attackers, the data in FDAA is often disguised, which makes the false data easily misjudged by the GRU prediction model as real data, which leads to the detection accuracy of the trained GRU prediction model for FDAA. are greatly restricted. Contents of the invention [0003] The technical problem to be solved...

AI Technical Summary

Problems solved by technology

However, for some experienced attackers, the data in FDAA is often disguised, which makes the false data easily misjudged by the GRU prediction model as real data, which leads to the detection accuracy of the trained GRU prediction model for FDAA. heavily constrained

Images