For towering structures under complex time-varying excitation, the tuned magnetic fluid rolling-ball dampers (TMFRBD) cannot achieve self-tuning damping. In this paper, considering the operating characteristics of the TMFRBD, a damper self-tuning control algorithm, namely particle swarm optimization (PSO)-FUZZY-PID, is proposed for real-time adjustment of the intrinsic frequency of TMFRBD, which can greatly improve the damping effect, namely self-tuning magnetic fluid rolling-ball dampers (STMFRBD). Firstly, the equivalent dynamics model of the damping system is established, deriving the structural dynamic amplitude response as a function of the excitation frequency to structural frequency ratio. Different frequency tracking schemes of the damper are compared and analyzed. Then magnetic field simulations and magnetic fluid magnetic property measurements are performed for the dampers, respectively. Secondly, the self-tuning PSO-FUZZY-PID control algorithm is designed to improve the damping performance of STMFRBD under time-varying excitation by optimizing factors of the FUZZY-PID controller with PSO. To confirm the effectiveness of the damper self-tuning control algorithm, the frequency tracking of dampers with different control algorithms under time-varying excitation is compared. The results show that the damper’s frequency can be tuned quickly and accurately tracking excitation frequency by PSO-FUZZY-PID. Finally, the simulation and experimental results of STMFRBD and passive mass damper (PMD) under different loads are analyzed, and the structural amplitude response is analyzed using the Hilbert-Huang transform (HHT). In addition, the damping of STMFBD and TMFBD under different loads is compared. The results show that STMFRBD has better vibration-damping performance than PMD and TMFRBD.