A prescribed-time fault-tolerant control method based on broad-learning neural network is proposed in order to deal with the actuator failure and uncertainty problems in helicopter systems. Firstly, a complete state model of the helicopter is established, and a prescribed-time stability control strategy is achieved by integrating prescribed-time control with dynamic surface control. Then, to handle actuator faults and model uncertainties, a broad-learning neural network is designed to estimate these perturbations. The estimation accuracy of the BLS is significantly improved through the incorporation of additional feature and enhancement nodes. Then, based on the estimated perturbations, compensatory control terms are constructed, leading to the development of an adaptive control law capable of effectively mitigating actuator faults and uncertainties. Finally, the effectiveness of the proposed approach is validated through simulation experiments, demonstrating its robustness and reliability.