Objective To investigate the accurate positioning and error compensation model of pneumatic actuators in food sorting machinery, and to improve positioning accuracy and dynamic response.Methods The effects of cylinder diameter, position sensor resolution, and friction force on positioning accuracy were analyzed. A correction method based on the Stribeck friction model was proposed, and the ARIMA model was employed to predict and compensate for pressure fluctuations. The control strategy was optimized by integrating a fuzzy PID self-tuning algorithm with a dual closed-loop control architecture.Results Validated through a digital twin platform, the maximum tracking error under sudden load changes was reduced from 0.38 mm to 0.05 mm, and the stabilization time was shortened to 0.48 s. In a continuous 24-hour experiment, the standard deviation of the positioning error remained within ±0.03 mm. During industrial validation, the positioning time was reduced from 120 ms to 85 ms, and the defect rate dropped to 0.15%.Conclusion The proposed model significantly improves the positioning accuracy, dynamic response, and stability of food sorting machinery.