In order to improve the torque tracking accuracy of a load simulator for a servo system, a method of composite terminal sliding mode control based on the fuzzy RBF neural network is proposed.Firstly, by analyzing the system's composition and working principle of the servo load simulator, the torque motor model is simplified.According to the models of the torque sensor and the inertia disk, the simplified equivalent model of the servo load simulator is established.Then, a fast terminal sliding mode controller is designed.In order to improve the dynamic quality of the sliding mode, the parameters of the sliding mode surface are dynamically adjusted by using the method of the fuzzy neural network.At the same time, in order to improve the learning and training speed of the fuzzy RBF neural network, the nearest-neighbor hierarchical clustering and the conjugate gradient algorithm are used to adjust the parameters of the network, and local optimization is carried out to improve the performance of the algorithm.Finally, simulations are conducted, and the results show that the method can improve the system's control accuracy and has good dynamic characteristics.