Objective: Solve the problem of low accuracy of collaborative robots in food dynamic target grasping at present. Methods: Based on the architecture of collaborative robots, a combination of fuzzy self-tuning PID control and robust adaptive compensator was proposed for collaborative robot food dynamic target grasping. PID combined with fuzzy control to achieve parameter self-tuning, and robust algorithm combined with adaptive algorithm for system uncertainty compensation. The performance of the proposed method was analyzed through experiments, verifying its feasibility. Results: The proposed method had good results in the dynamic target grasping of collaborative robots, improving the accuracy of dynamic grasping. At a conveyor belt speed of 100 mm/s, the success rate of dynamic grasping reaches 99.50%, which had certain application value for food dynamic target grasping. Conclusion: By optimizing existing target grasping control methods, the grasping accuracy of collaborative robots can be effectively improved.