基于改进卡尔曼滤波的CPS虚假数据注入攻击检测

CPS False Data Injection Attack Detection Based on Improved Kalman filter

为提高信息物理系统(Cyber-Physical Systems,CPS)安全,提出一种改进卡尔曼滤波(Kalman Filtering,KF)的CPS虚假数据注入攻击(False data injection attack,FDIA)检测方法。首先,在标准KF中引入渐消记忆指数加权方法,以使KF算法自适应统计噪声特性,从而提高滤波性能;然后利用改进KF算法构建攻击检测模型,对FDIA进行检测;最后,在IEEE-30节点和IEEE-14节点系统上进行仿真,验证改进KF算法的有效性。结果表明,改进KF算法可有效检测FDIA,且相较于标准KF算法、加权最小二乘滤波(weighted least squares,WLS)算法,改进KF算法对FDIA检测结果与实际值更为接近,平均绝对误差分别为2.1336和1.2543。由此表明,本改进算法可用于CPS的FDIA检测。

In order to improve the security of information Physical system(CPS),a detection method of CPS false data injection attack with improved Kalman filter(KF)is proposed.Firstly,the gradual fading memory index weighting method is introduced into the standard KF to make the KF algorithm adapt to statistical noise characteristics and improve filtering performance;Then,an attack detection model is constructed using the improved KF algorithm to detect CPS false data injection attacks;Finally,simulations were conducted on IEEE-30 node and IEEE-14 node systems to verify the effectiveness of the improved KF algorithm.The results show that the improved KF algorithm can effectively detect CPS false data injection attacks,and compared with the standard KF algorithm and WLS algorithm,the detection results of the improved KF algorithm for false data injection attacks are closer to the actual values,with Mean absolute error of 2.1336 and 1.2543 respectively.This indicates that this improved algorithm can be used for attack detection of CPS false data injection.