A drag teaching spraying robot based on process closed-loop, which achieves adaptive human-machine drag teaching through force position hybrid control. By constructing a real-time solution framework for multi degree of freedom dynamic models to automatically identify system modes, combined with torque closed-loop control, compliant force tracking characteristics and accurate spray trajectory motion characteristics are achieved. During the demonstration process, the system achieves a balance between gravity and friction based on model identification, while recognizing the demonstration intention using a drag state classifier based on support vector machines. This improves the flexibility and dexterity of the operation, distinguishes between operating force and accidental collisions, and ensures the generation of trajectories that meet the requirements of the spraying process. In the process of trajectory reproduction and spraying processing, accurately reproduce the process flow according to the taught trajectory and recorded time sequence, and achieve the processing goals. At the same time, based on the joint modeling of fluid dynamics and kinematics, combined with the input of machining effects, self optimization of process parameters and motion parameters is achieved, thus realizing closed-loop control of the process.