拒绝服务攻击下基于UKF的智能电网动态状态估计研究

针对连续拒绝服务(Denial of service, DoS)攻击导致量测数据丢失使得动态状态估计失效、进而破坏智能电网安全经济运行问题,本文提出了一种适用拒绝服务攻击的改进无迹卡尔曼滤波(Unscented Kalman filter, UKF)方法,以进行智能电网动态状态估计.首先,分析拒绝服务攻击引起数据丢包特性并设计了数据补偿策略,以重构电力系统动态模型;然后,结合Holt s双参数指数平滑和无迹卡尔曼滤波方法,构造了融合补偿信息的新状态估计方程,并进一步基于估计误差协方差矩阵推导了状态增益更新方法,从而得到了无迹卡尔曼滤波动态状态估计新方法.最后,针对IEEE 30和118节点系统进行仿真,验证了所提方法的可行性和有效性.

Dynamic State Estimation of Smart Grid Based on UKF Under Denial of Service Attacks

When continuous denial of service (DoS) attacks cause measurement data losses in smart grid, the traditional dynamic state estimation is useless, destroying the running safety of smart grid seriously. To solve the problem, an improved unscented Kalman filter (UKF) is proposed, which can estimate the dynamic state of smart grid under DoS attacks. Firstly, the characteristics of data packet losses resulting from DoS attacks are analyzed and data compensation strategy is designed to reconstruct the dynamic model of power system. Integrating Holt's two-parameter exponential smoothing and unscented Kalman filter algorithms, a new state estimation equation including the compensation information is then constructed. Furthermore, a state gain updating method is derived from the estimated error covariance matrix, which produces a new enhanced UKF dynamic state estimation algorithm. Finally, simulations on IEEE 30-bus and 118-bus system confirm the feasibility and effectiveness of the proposed method.