Operating-Condition-Dependent Feedforward Control Strategy for Primary Frequency Regulation of Hydropower Units

In the context of building a novel power system and with the increasing penetration of renewable energy, hydropower units are increasingly required to operate under wide-range operation (WRO) conditions. However, under the opening mode, the unit’s primary frequency regulation (PFR) performance is significantly affected by head and load fluctuations. This poses a risk of failing grid assessment requirements. To enhance the PFR performance, this study first establishes a nonlinear simulation model of the hydro-turbine regulation system (HTRS) for PFR and conducts a simulation analysis of PFR performance under varying head, load, and frequency deviation conditions. Subsequently, based on the simulation results, an operating-condition-dependent feedforward–feedback control strategy is proposed. This strategy utilizes a BP neural network (BPNN) to establish the head-power-opening (H-P-Y) mapping relationship, calculates the feedforward opening command in real time, and superimposes it with the PID feedback correction to form the total guide vane opening (GVO) setpoint. Performance comparisons through multi-condition simulations within the ranges of 70–100 m head, 10–90%Pr load, and 0.05–0.15 Hz frequency deviation demonstrate that the proposed strategy ensures that all selected assessment indices meet PFR compliance standards across all tested conditions. This study provides a feasible pathway for improving the PFR performance of hydropower units under WRO in the opening mode.