<p>Lower limb rehabilitation robots often assist patients using predetermined gait trajectories, whose control accuracy is susceptible to model uncertainties and parameter perturbations. To address this, this paper proposes a motion control strategy that combines dynamic movement primitives (DMPs) with fractional-order recursive terminal sliding mode control. The aim is to overcome the limitations of traditional methods, such as their reliance on fixed trajectories and insufficient disturbance rejection capability. First, based on demonstrated trajectories from healthy subjects, DMPs are employed to generalize the amplitude and frequency of rehabilitation training trajectories through nonlinear superposition of basis functions and phase coupling, thereby generating desired reference gait trajectories. Subsequently, a fractional-order recursive terminal sliding mode control algorithm is designed by integrating Riemann-Liouville fractional-order theory with sliding mode control and incorporating a recursive error compensation structure, ensuring rapid convergence of tracking errors. Finally, the global fixed-time convergence of the closed-loop control system is rigorously proven based on the Lyapunov stability criterion. The effectiveness and superiority of the proposed method are validated through experiments on a lower limb rehabilitation robot platform.</p>