<p>This study addresses the disturbance limitations such as sudden load changes and parameter perturbations of computer numerical controller (CNC) machines driven by permanent magnet linear synchronous motors (PMLSMs), which can lead to decrease accuracy and weakening of system robustness. To overcome these challenges, we propose a fractional-order terminal sliding mode control (FTSMC) based on model reference adaptive system (FTSMC-MRAS) for PMLSM drives. In this strategy, sudden load changes are integrated into the mover mass. The discrete MRAS identification theory is employed to identify the load mover mass. Based on this identification, the compensation current is calculated and applied as feedforward compensation to mitigate disturbance effects, thereby enhancing the system global robustness. An FTSMC is designed to suppress the impact of system uncertainties, ensuring that the system converges within a finite time frame. To overcome the inherent chattering problem associated with SMC and improve system tracking performance, a power-exponential adaptive switching law is designed. This approach avoids excessively high control gains that could compromise system stability, enabling variable-speed approaching the sliding mode zero-value neighborhood. Consequently, it ensures rapid state convergence while mitigating chattering effects. Finally, we conduct experiments under different operating conditions to validate the proposed control strategy.</p>