Abstract: In order to improve the driving control accuracy of the micro-motion part of the macro-micro composite actuator, a composite control strategy based on long short-term memory (LSTM) -fuzzy PID is proposed. The structure and working principle of the macro-micro composite actuator are briefly described, and the mathematical model of the micro-motion part is established based on the principle of magnetostrictive. An LSTM neural network inverse model with higher adaptability to experimental data was trained, and a fuzzy PID controller was introduced to build an inverse model displacement compensation control system for the micro-motion part. Through numerical simulation and experiment, the superiority of LSTM-fuzzy PID composite control strategy is verified. The experimental results show that the maximum error of LSTM-fuzzy PID composite control is less than 1.7% under the sine wave signals with frequencies of 1, 5 and 10 Hz, respectively. Compared with LSTM-PID composite control, the maximum error is reduced by 69%, 68% and 68%, respectively. The research results show that the LSTM-fuzzy PID composite control strategy has higher positioning accuracy and smaller overshoot, which can better meet the micro-motion control requirements of macro-micro composite actuators.