信息物理系统降阶观测器设计及传感器攻击检测

Reduced-Order Observer Design and Sensor Attack Detection of Cyber-Physical Systems

提出了一种适用于含有未知输入信号的信息物理系统降阶观测器设计方法及传感器攻击检测应用.首先,利用克罗内克积将信息物理系统描述为整体系统动态模型;然后考虑系统存在未知输入时,通过增加降阶观测器增益矩阵的约束条件,实现降阶的同时消除未知输入给系统带来的影响,再通过系统输出状态判断是否存在传感器攻击信号.最后,通过数值仿真,验证了所提降阶观测器对系统的状态估计及检测传感器攻击信号的能力,且对比仿真结果显示,所提降阶观测器性能优于全阶观测器.

In this paper,a design method of a reduced-order observer for the cyber-physical system with unknown input signals and its application in sensor attack detection are investigated.Firstly,the Kronecker product is used to describe the original cyber-physical system as a dynamic model of the whole system.Then,a reduced-order observer with unknown input for cyber-physical system is presented,and the influence of unknown input on the system can be eliminated by increasing the constraint condition of the reduced-order observer gain matrix,and then judging whether there is a sensor attack signal through system output status.Finally,the capability of the proposed reduced-order observer to estimate the state of the system and detect the attack signal of the sensor is verified by numerical simulation.Also,the simulation results show that the performance of the proposed reduced-order observer is better than that of the full-order observer.