In order to improve the loading accuracy of certain type servo system loading simulator, a gray prediction single neuron PID control strategy is designed. By analyzing the system configuration and working principle of the servo system loading simulator, the model of the load motor is simplified, and a servo load simulator model is established according to the inertia moment disk model. Based on the traditional PID control, a gray prediction model is added to initialize the PID parameter setting, and a single neuron adaptive controller is added to adjust the PID proportional, integral and derivative parameters online. The simulation results show that the designed control method can improve the loading accuracy of the loading system, and it has strong robustness and is superior than the traditional PID control.