<p>This study focuses on addressing the issue where the permanent magnet synchronous motor (PMSM) servo system is prone to being affected by uncertainties like load disturbances, resulting in subpar speed tracking accuracy. The core challenge of this problem lies in the fact that conventional sliding mode control, due to its discontinuous switching, tends to induce high-frequency chattering in the control input current, which not only degrades actuator lifespan but also adversely affects disturbance estimation. Meanwhile, existing disturbance observers typically neglect the coupling between chattering and estimation error, making it difficult to achieve accurate disturbance reconstruction under highly dynamic operating conditions. To address these challenges, a control strategy integrating the super-twisting sliding mode controller (STSMC) and an improved disturbance observer (IDOB) is developed. Firstly, the STSMC along with its switching control law is designed to eliminate discontinuities, thereby reducing system chattering and enhancing the system’s tracking precision. Secondly, with the aim of estimating load disturbances in the system, the IDOB is employed to feed forward the estimated load disturbance values to the speed controller for compensation. This serves to suppress chattering in the control input current and strengthen the system’s robustness. Considering the problem of a large number of parameters involved, the particle swarm optimization (PSO) algorithm is utilized to optimize the controller parameters, obtaining the optimal parameter configuration for the controller. This significantly boosts the transient response speed and substantially enhances the system’s robustness. Finally, experimental results verify that the proposed control method achieves excellent tracking accuracy, strong robustness, and a fast transient response speed.</p>