A hierarchical clustering-based optimization strategy for active power dispatch of large-scale wind farm

For large-scale wind farm, the active power dispatch strategy should balance the conflicts among the tracking accuracy, regulation flexibility and solver reliability, which are not well achieved by the conventional proportional distribution (PD) strategy. In this paper, a novel hierarchical-active-power-dispatch strategy is proposed for the larger-scale wind farm based on the fuzzy c-means clustering algorithm and model predictive control method. Firstly, both the power tracking dynamic characteristics and output power fluctuations of wind turbines are considered as decision variables to divide the wind farm into appropriate clusters. Then the wind farm active power dispatch strategy can be constructed within a hierarchical control framework. More concretely, a lower-layer proportional controller is synthesized with the conventional PD strategy to distribute the active power for wind turbines within a cluster, which forms a closed-loop structure with robustness. The MPC strategy is adopted in the upper layer to dispatch the active power control set-point from the wind farm-level to clusters, which has fully considered the dynamic characteristics of each cluster. The proposed hierarchical strategy has the advantages of reducing the optimization problem scale, eliminating the dynamic tracking errors, enhancing the dynamic dispatching stability and robustness and increasing the active power distribution flexibility. Simulation results show the significant improvement and good robustness of the proposed strategy.