[Objective] Permanent magnet synchronous motor (PMSM) is widely adopted in industrial applications due to their high power density, superior efficiency, and excellent dynamic performance. However, conventional control strategies, often exhibit degraded performance when confronted with system parameter variations and external disturbances. To address these challenges, this paper proposes a novel sliding mode variable structure control (SMVSC) strategy optimized by the marine predators algorithm (MPA), aiming to enhance system robustness and steady-state accuracy. [Methods] Within the vector control framework of PMSM, the conventional proportional integral controller in the speed outer loop was replaced with SMVSC controller. The inherent robustness of SMVSC was leveraged to mitigate the adverse effects of parameter uncertainties and external disturbances. To further refine the SMVSC design, a double-power reaching law was introduced to optimize the approaching process. During the initial phase of converging to the sliding mode surface, a high gain was employed to accelerate convergence speed. Subsequently, the gain was dynamically reduced in the later phase to suppress chattering phenomena, thereby balancing rapid dynamic response with control smoothness. Additionally, the MPA was utilized to optimize critical parameters of the SMVSC, including reaching rate coefficients and sliding surface parameters, avoided local optima and ensures the adaptability of controller parameters under diverse operating conditions. Finally, the simulation model of PMSM speed regulation system was built by Matlab/Simulink, and experiments were carried out on the physical platform, verified the effectiveness of the proposed method. [Results] Simulation and experimental results showed that the MPA-optimized double-power reaching law SMVSC significantly improves PMSM speed regulation performance, with enhanced stability, reduced overshoot, and shorter settling time. Compared to traditional methods, the optimized controller exhibits superior dynamic response and robustness under varying conditions. [Conclusion] The proposed MPA-optimized SMVSC strategy effectively enhances PMSM control performance under parameter uncertainties and external disturbances, offering a new solution for high-performance of PMSM.