Composite Anti‐Disturbance Control of Networked Stochastic Systems Subjected to Bias Injection Attacks

ABSTRACT The networked stochastic systems (NSSs) affected by disturbances and bias injection attacks (BIAs) are studied, which are representative in practical engineering applications. To mitigate the negative impact of disturbances and BIAs on the NSSs, the interconnected observers include stochastic disturbance observer (SDO) and a stochastic attack observer (SAO), have been designed, which can effectively leverage the available information. In a more general sense, the additional attack information is generated when the BIA occurs, and then the neural network (NN) method is used to approximate it by selecting an appropriate neural network basis function. On this basis, the compound observer‐based controller is constructed to compensate for the disturbances and BIAs, ensuring that the systems can converge to a bounded region. Finally, the proposed scheme is demonstrated to be effective through simulation experiments.