Event-triggered neural adaptive backstepping control of the K chaotic PMSGs coupled system

Abstract This paper presents an event-triggered neural adaptative backstepping control method for the K chaotic permanent magnet synchronous generators (PMSGs) coupled system. The actual wind farm is usually a complex coupled system, in which there are K identical generators operating simultaneously and forming a coupled electromagnetic field through the transmission bus. By using a series of capacitors and resistors to concatenate multiple identical PMSGs forming a coupled network, we firstly establish its dynamic model based on Faraday's law of electromagnetic induction. Then we conduct its dynamic analysis and find this coupled system can generate complex nonlinear dynamics like chaos oscillation, which, if not being compensated properly, could lead to severe deterioration of system performance. Here we propose an adaptive backstepping scheme by integrating the simplified interval type-2 fuzzy neural network (SIT2FNN), Nussbaum type function, improved saturation function reaching law, cosine barrier function, event-triggered strategy and tracking differentiator (TD). In the setting, the SIT2FNN is utilized to estimate unknown terms of the system and the improved reaching law is used to improve the stability of the controller along with the Nussbaum type function to approximate unknown parameters. Meanwhile, the event-triggered strategy is exploited to save computation and communication resources along with the second-order TD to avoid the term explosion associated with backstepping. The stability of the proposed scheme is proved by the Lyapunov function. Finally, the effectiveness of our scheme is verified by simulation results.