Efficient Distributed MPC Using Kautz Functions for Load Frequency Control in Multi-Area Power Systems With Communication Delays

The rapid increase in renewable energy deployments and its integration into power grids bring challenges in balancing supply and demand, leading to frequency deviations. Consequently, various load frequency control (LFC) strategies have been used to ensure frequency stability. However, the centralized strategies exhibit limitations in efficiently responding to the evolving conditions of distributed energy sources. To address these challenges, this paper proposes a distributed model predictive control (MPC)-based method for LFC of multi-area power systems. Discrete Kautz functions, featuring flexible non-identical pole configuration, are utilized to approximate the predictive control trajectory, reducing MPC's computational complexity. The proposed Kautz functions-based distributed discrete MPC (KDD-MPC) effectively handles various nonlinear constraints in LFC and integrates information from neighboring areas as feedback to realize inter-area coordination. A delay compensation mechanism is implemented to improve robustness under time-varying communication delays. The performance of this LFC approach is evaluated through numerical experiments on the IEEE 39-bus system, showing a 10.8% reduction in the mean absolute percentage area control error (MAPACE) in the disturbed area with wind power participation, compared to the PSO-optimized PI controller. Meanwhile, the computational time is reduced by 83.6% relative to conventional MPC, with the prediction and control horizons set to 25.