<p>This paper proposes a prescribed-time convergent adaptive fault-tolerant iterative learning control method for nonlinearly parameterized repetitive motion systems subject to actuator faults and input saturation constraints. A novel nonlinear error transformation function that integrates sublinear and superlinear terms is designed, and an accelerated transformation function is employed to convert the tracking problem under faulty conditions into a convergence problem of the transformed error. The transformation function is identically zero at the initial instant, which completely decouples the convergence time from the system initial conditions. Based on this transformed error a new PD-type fault-tolerant control law is constructed. A normalized adaptive update law is developed to estimate the upper bound of the lumped uncertainty online, while an iterative learning feedforward term is incorporated to progressively learn the repetitive component of the fault across successive iterations, thereby achieving coordinated compensation for multiplicative efficiency loss, additive bias faults, and time-varying parametric perturbations. Simulation results on a DC motor servo system demonstrate that, under conditions where the actuator efficiency drops to 70% with periodic additive faults, the tracking error converges to a neighborhood of zero within the prescribed time. After 20 iterations, the interval-wide RMS error is significantly reduced and the control input satisfies the saturation constraint. Compared with methods lacking a fault-tolerant mechanism, the proposed approach exhibits marked improvements in both steady-state tracking accuracy and iterative convergence rate under faulty conditions.</p>