Mode identification-based model-free adaptive predictive damping control method for power system with wind farm considering communication delays

Abstract

With the widespread utilization of large-scale power electronics and the integration of wind power into power grid, the problem of low-frequency oscillations in new power systems become increasingly critical. In this paper, a mode identification-based model-free adaptive predictive control (MI-MFAPC) method with full-form dynamic linearization (FFDL) is proposed to design wide-area damping controller (WADC) to suppress multi-modal oscillations. Modal identification method is adopted in MFAPC-WADC to screen the controller input signals with high observability and low coupling, and the controller parameters are set separately for different oscillation modes. The improved control and adaptive law are proposed to rigorously guarantees the asymptotic convergence of the control error to zero under steady-state conditions. The proposed method integrates with the rolling optimization and introduces the future time input and output data, so it has strong robustness for the time-delay system. Simulation results show that MI-MFAPC-WADC greatly improves the dynamic quality and system stability, and the multiple oscillation modes are suppressed efficiently. Considering communication delays, the proposed method without the adaptive delay compensator even obtains better damping performance than the existed method with the adaptive delay compensator.