The work aims to design an intelligent filling control system for packaging production line to improve the filling precision and speed. The basic structure of filling production line, including delivery, empty bottle detection, filling, quality detection, secondary filling, cover and finished product output stations, was described. Taking the secondary filling irrigation as the research object, a micro filling control system based on PID control of cerebellum neural network was designed, which could make up for the deficiency of conventional PID control. At the same time, kalman filter was used to eliminate the interference signal and noise during operation. The feasibility and effectiveness of the control system were verified by experiments. The filling characteristics were analyzed and there were many factors affecting the filling accuracy which interfered with each other. The experimental results showed that the quality deviation of all samples did not exceed 0.1 g, and the production line ran stably, quickly and reliably during the filling process. The micro filling control system can improve the filling accuracy and speed, and can meet the requirements of high precision and high speed of packaging production line.