Abstract: In view of problem that traditional analysis method is difficult to obtain dynamic feature when mine hoist system happened fault, the paper proposed a self-tuning fault-tolerant PID control method for mine hoist based on BP neural network. The method employs neural network to predict system output value by learning and tracking dynamic features of the hoist system online, uses self-tuning neural network PID of self-adapt control to build fault-tolerant controller, so as to realize stable fault-tolerant control of the hoist system when fault happens. The simulation result shows that control method can track system fault state rapidly when a sudden fault happens, and adjust PID parameters online and restore system features in short time.