Adaptive Fuzzy Detector-Based Secure Correct Control for Cyber-Physical Systems subject to Heterogeneous Physical Attacks

Abstract

In this paper, the secure correct control problem is investigated for Cyber-Physical Systems (CPSs) subject to hetero-geneous physical attacks which include actuator attacks and sensor attacks. In order to better fit the actual attack environment, the actuator attacks are modeled as a class of completely unknown stochastic signals and the completely unknown nonlinear attack signals, the sensor attacks are modeled as a class of energy-limited stochastic signals that can realize coupling with s-tates to invade the sensor by wireless network, aiming at tampering with the state information of CPSs. An adaptive stochastic attacks detector (ASAD) is constructed to online detect the stochastic attack signals acting on actuator. As for the completely unknown nonlinear attack signals, the fuzzy approximation technology is introduced to correct the actuator signals. An adaptive fuzzy secure correct control (AFSCC) scheme is proposed by integrating ASAD with fuzzy approximation technology and robust method. It is proved that the security control performance of the CPSs is realized. Finally, a simulation examples is given to shows the effectiveness of the proposed scheme.