The conventional steam valve predominantly manages power regulation tasks while also addresses limited-scale primary frequency regulation. Condensate throttling can change energy distribution of the unit to a certain extent, and serve as a potential and selectable auxiliary frequency regulation method. To delve deeper into the frequency regulation capability of condensate throttling, the working principle and advantages are dissected, and static and dynamic models of the condensate throttling system are established, the frequency regulation characteristics are analyzed and the frequency modulation boundary conditions are outlined. To comprehensively enhance the dynamic performance of the condensate throttling primary frequency regulation, the system dynamic model is linearized, and then a model-switching fuzzy predictive control strategy is proposed. In this control strategy, fuzzy logic is introduced into the predictive controller algorithm to dynamically adjust control weighting coefficients in real-time, and predictive models are dynamically switched according to operational changes to enhance control quality. Case studies based on actual data from a certain 600 MW unit are conducted, indicating the static and dynamic frequency modulation capabilities of condensate throttling increase with the unit load, which has engineering application value in primary frequency regulation. Compared with the conventional PID and self-tuning fuzzy parameter PID controllers, the proposed control strategy exhibits superior adjustment time and performance metrics under various operating conditions, demonstrating better adaptability to changes in operating conditions.