Fdias Identification Of Distribution Network Based On Pseudo Measurement Modeling And Dynamic State Estimation

The increasingly close coupling interaction between information network and physical network makes the power system show the characteristics of the typical cyber physical system(CPS).But at the same time,it also facilitates the cyber hackers to transfer the risk of the information side to the physical side,resulting in the failure of power equipment,blackout accidents and other serious consequences.As a new type of cyber-attack,false data injection attacks(FDIAs)can tamper with the state estimation results,manipulate the power flows when avoiding the bad data detection and identification,which seriously threaten the safe and stable operation of power system.In addition,the data protection ability of a variety of remote terminal equipment in the distribution network is weaker than transmission network,the computing and information processing capacity is also limited,which makes the distribution automation system faced with greater cyber-attack threats.This paper firstly introduces the definition,classification and research status of power system state estimation,and then expounds the loophole of the algorithm for detecting and identifying the bad measurements in the energy management system(EMS).On this basis,we introduce the launching principle of the FDIAs and the attack constructed strategies under two information conditions.Considering the shortage of measurement redundancy in distribution network,a pseudo-measurement modeling method based on cloud adaptive particle swarm optimize spiking neural network(CAPSO-SNN)is proposed.Furthermore,the fusion estimation results of extended Kalman filter(EKF)and unscented Kalman filter(UKF)are taken as the prior suggested distribution of particle filter(PF)to construct a mixed Kalman particle filter(MKPF)based dynamic state estimation algorithm for distribution network.Finally,an attack identification method for FDIAs in distribution network based on global and single node variable test is proposed according to the dynamic hysteresis characteristics of the nonlinear filter.The proposed FDIAs identification method is verified in detail according to the simulation cases of IEEE-33 bus distribution system.The results show that the proposed pseudomeasurement modeling method can output high-precision pseudo-measurement data and improve the measurement redundancy of the system.In addition,the proposed MKPF algorithm exhibits more stable and accurate dynamic state estimation performance.Finally,two typical FDIAs examples verify the validity and practicability of the proposed identification method.