[Objective] Ultra-low frequency oscillations(ULFOs)

typically occurring in the range of 0.1～0.7 Hz

have become a critical stability issue in power systems with a high penetration of hydropower units. In practical operation

inappropriate or poorly tuned governor control parameters may introduce insufficient damping or even negative damping

thereby triggering or amplifying ULFOs. The objective of this study is to develop an effective and engineering-oriented parameter optimization method for hydropower unit governors to enhance low-frequency damping performance and improve dynamic stability under disturbances. [Methods] A governor parameter optimization framework based on an improved particle swarm optimization(PSO) algorithm is proposed. The method is established on a typical generator–turbine coupled model that captures the hydro-mechanical dynamics of the water diversion system

turbine

and generator. A fractional-order PID(FOPID) controller is adopted to provide higher flexibility and control accuracy compared with conventional integer-order PID controllers. To overcome the limitations of traditional PSO

such as premature convergence and weak global search capability in complex multi-parameter optimization problems

several improvements are introduced. First

a multi-strategy particle initialization scheme is employed to enhance the uniformity and diversity of the initial population

providing a more favorable starting point for global exploration. Second

a dynamic diversity enhancement mechanism is developed to monitor population evolution in real time and adaptively adjust the search strategy

thereby preventing stagnation and maintaining exploration ability during iterations. Third

a robust fitness function is constructed by comprehensively integrating key performance indicators

including low-frequency oscillation characteristics

regulation time

overshoot

and steady-state error. These indices are jointly considered to form a multi-objective optimization framework specifically oriented toward low-frequency performance and engineering feasibility. [Results] Comparative simulation studies are carried out to evaluate the effectiveness of the proposed method. The improved PSO-based approach is systematically compared with traditional PSO

enhanced PSO

the whale optimization algorithm(WOA)

the grey wolf optimizer(GWO)

and an empirical tuning method. Simulation results demonstrate that the proposed method exhibits clear advantages in suppressing ultra-low frequency oscillations. Specifically

the optimized controller significantly reduces oscillation amplitude and damping time

shortens system regulation time

and effectively suppresses response overshoot. Moreover

under external disturbances and parameter variations

the optimized governor shows improved dynamic stability and stronger parameter robustness compared with the benchmark methods. The results indicate that the proposed improved PSO achieves a better balance between convergence efficiency and global optimality

leading to superior overall control performance. [Conclusion] The proposed improved PSO-based governor parameter optimization method effectively enhances low-frequency damping characteristics and mitigates ultra-low frequency oscillations in hydropower units. By combining multi-strategy initialization

dynamic diversity enhancement

and a robust fitness function design

the method achieves reliable and high-quality parameter tuning without modifying hardware structures or higher-level control architectures. The study provides a practical and scalable technical solution for oscillation suppression in hydropower-dominated power systems and offers valuable guidance for engineering applications involving governor parameter tuning under complex operating conditions.