Xin Wang

This work addresses the reinforcement learning (RL)-based dynamic event-triggered control problem for Markov jumping power systems (MJPS) subject to transient line faults and denial-of-service (DoS) attacks. First, to conserve communication resources and control costs, a novel RL-based dynamic event-triggered protocol (DETP) is established, which models the control policy and the event-triggered error as zero-sum game players, with the aim of obtaining the worst event-triggered error to maximize the triggering interval. Meanwhile, from the viewpoint of the defender, we incorporate the impact of DoS attacks into the performance value function, which enhances the robustness of the learned control policy and ensures the system’s reliable operation in hostile environments. Then, to obtain the Nash equilibrium solution of the secure optimal control problem (OCP), a data-driven improved policy iteration (IPI) algorithm is introduced without relying on the system dynamics. By incorporating a constant sequence, the algorithm not only avoids to seek the initial stabilizing control policies (ISCPs) and improves learning efficiency through faster convergence compared to traditional PI and value iteration (VI) methods. Finally, this paper provides the theoretical guarantees on convergence and stability of the proposed algorithm, and simulation results also validate the effectiveness and superiority of the proposed approach.

Due to the open communication architecture and the ever-expanding network connections of modern power systems, cybersecurity has become increasingly important. For this reason, this paper tackles the load frequency control (LFC) problem in multi-area power systems (MAPSs) that are vulnerable to both time-constrained denial-of-service (DoS) attacks and unknown false-data-injection (FDI) attacks. First, unlike existing research that relies on randomly generated DoS attacks, this paper proposes a detection algorithm aimed at identifying DoS attacks based on network traffic analysis. Moreover, to address unknown FDI attack, we employ neural networks (NNs) to learn the characteristics of the attack signals and construct an approximation model. Then, an NN-based attack compensation controller is developed to mitigate or eliminate the negative impact of unknown maliciously FDI information on system performance. By carefully constructing Lyapunov functional, the exponential asymptotically stability condition with an ${\mathcal {H}}_{\infty }$ performance index is derived under hybrid attacks. We also determined the relationship between the performance index of the compromised LFC system and the minimum allowable sleep time and maximum allowable active time under DoS attacks. Finally, a simulation example is illustrated to substantiate the validity of the proposed approach.

This article studies the input-to-state stability (ISS) problem of networked control systems (NCSs) subject to both Denial-of-Service (DoS) and unknown deception attacks (DAs). A neural network (NN)-based resilient event-triggered control protocol (RETCP) is first presented to mitigate resource constraints and the adverse effects of cyber attacks, where the NN technology is leveraged to neutralize and approximate the malicious data injected by unknown DAs. Then, we develop a new predictor to compensate for lost signals of NCSs during the DoS threats, so that the NCSs can further tolerate more unfavorable DoS and unknown DAs. It is shown that the resulting NCSs with the designed novel NN-based controller can achieve ISS under the complex attacks. Finally, experimental evaluations are conducted for an uncrewed ground vehicle (UGV) to verify efficacy of the proposed intelligent control protocols.