Abstract: In order to ensure the safe and stable operation of the hydropower plant units and optimize the maintenance strategy, this paper proposes a prediction algorithm for the remaining life of the head cover drainage pump for the water turbine head cover drainage system. On the basis of real-time data, combined with the reliability principle, the corresponding operation reliability of the head cover drainage pump is calculated according to its vibration speed, vibration displacement, pumping flow and pumping time, and the integrated operation reliability of the head cover drainage pump is obtained. Combined with the limit learning machine algorithm, the residual life assessment model is constructed, and the dynamic residual life of the top cover drainage pump of the hydraulic turbine is comprehensively predicted. Combined with the real-time data of the top cover drainage system of the hydropower plant, the effectiveness and feasibility of this method are verified.