Resilient supervisory control under actuator attack with unknown energy

This paper delves into the design issue of resilient supervisory control in Discrete Event Systems (DESs) under actuator-enabled attacks. More precisely, the model abstracts event weights as the energy costs for the attacker, which enables the actuator attacks to be captured even when the attacker’s energy remains unknown. Using supervisory control designs, this approach aims to prevent attackers from expending sufficient energy to carry out attacks on disabled events, thus avoiding unsafe states. Initially, by quantitatively computing the minimum and maximum critical thresholds, the actuator attack model with unknown energy is characterized as three different energy intervals. And the necessary and sufficient conditions are explored for the existence of a resilient supervisor under this actuator attack model within different energy intervals. Subsequently, an algorithm to implement this novel attack structure, namely the All Actuator-Enabled Attack (AAEA) structure, is developed. It is intended to grasp the relationship between the supervisor and the environment, so as to obtain an attack-resilient supervisor along with its corresponding control strategy. Finally, an illustrative example is presented to validate the effectiveness of the proposed method.