面向非高斯噪声干扰和拒绝服务攻击下的电力系统状态估计方法

With the gradual development of the traditional power grid into a power cyber-physical system, non-Gaussian noise interference and random denial of service (DoS) attacks will inevitably occur, which will lead to the low estimation accuracy of the traditional Kalman filter algorithm in power system state estimation. In this paper, the DoS attack compensation strategy is used to reconstruct the power system model, and the Cauchy kernel minimum error entropy cubature Kalman filter (CKMEE-CKF) algorithm is proposed to estimate the dynamic state of the power system. The proposed algorithm is based on the augmented model constructed by the statistical linearization method. The minimum error entropy (MEE) is used as the optimal criterion, and the state error and measurement error are combined into the MEE cost function. At the same time, the Gaussian kernel function in MEE is replaced by the Cauchy kernel, which is insensitive to kernel width, which greatly simplifies the difficulty of selecting kernel width and effectively avoids the singularity of Cholesky decomposition. Then, the fixed-point iteration algorithm is used to update the estimation recursively. Finally, in the IEEE-30 node system and the IEEE-118 node system, the CKMEE-CKF algorithm and the CKF and MEE-CKF algorithms, respectively, are used to estimate the state of the power system under various noise environments and DoS attacks. Taking the root-mean-square error of voltage amplitude estimation of the IEEE-30 node system as an example, compared with CKF and MEE-CKF algorithms, the experimental results show that the estimation accuracy of the new algorithm is improved by 88% and 60%, respectively, under the interference of the third non-Gaussian noise. Under the first DoS attack, the estimation accuracy is improved by 91% and 70%, respectively. In the case of non-Gaussian noise interference and DoS attack, the estimation accuracy of the new algorithm is significantly improved, and it is an effective method for power system state estimation.