Dynamic Event-Triggered Non-Vulnerable Fault-Tolerant Control for Wind Power Cyber-Physical Systems under Cyber-Attacks

Abstract

The study addresses the problem of the insufficient robustness of traditional control strategies under network attacks by quantifying the impact of attacks. It accomplishes this by establishing the dynamics of the wind turbine and the sensor/actuator attack model, designing a dynamic event trigger mechanism with an adaptive trigger threshold adjustment to reduce communication frequency, developing a fault-tolerant controller based on state feedback and fault compensation, and optimizing the control gain matrix using Lyapunov theory. The simulation results showed that, under Scene 3 (maximum attack intensity), the recovery time for rotor angular velocity was less than 4 seconds, communications are reduced by 70% compared to the traditional strategy, and energy consumption increases by no more than 25%. The sensor measurement error was reduced from 0.23 to 0.01. The comprehensive robustness index was 0.75, and the gain margin and phase margin were maintained at 4.5dB and 35.5° respectively. This research provides a control scheme for the physical system of a wind power network that balances security, real-time performance, and energy efficiency. It has engineering value in improving the ability of new energy power systems to resist interference.