Distributed Fusion Estimation for Nonlinear Cyber-Physical Systems With Attacked Control Signals

This article studies the secure estimation problem for cyber-physical systems where the control input is falsified by false data injection attack. First, novel optimization objectives are constructed and attack estimation is performed from real and expected measurement outputs, where the estimator gains are jointly derived through solving quadratic problems recursively. Subsequently, the exponential stability of the proposed estimator is derived if the nonlinear observability rank condition is satisfied, and an introduced factor is given properly. Moreover, to enhance estimation performance, the distributed fusion criterion is designed and the weighting matrices are obtained without any statistical information. Finally, an intelligent localization system monitored by multiple sensors is given to illustrate the effectiveness of the proposed method.