Floating electromagnetic constant force grinding and polishing actuator and its dynamic control

To address the issues of low efficiency, poor working environment, and unstable quality associated with manual polishing of freeform surface products, this paper proposes an end-effector constant-force actuator for robotic rigid contact polishing. To meet the high dynamic response requirements of high-speed polishing of freeform surfaces, the constant-force actuator is designed with a floating structure that includes a stator and a mover. The mover drives the polishing tool to maintain low inertia linear sliding along the stator guide rail. By applying a specific current to the mover coil, electromagnetic force is directly generated, and through precise control of the coil current, the contact force between the polishing process and the workpiece can be regulated in real-time, enabling compliant contact between the rigid abrasive and the workpiece. To enhance the stiffness during polishing and avoid frictional stick-slip, the actuator's mover is designed with a hollow structure, placing the polishing consumable clamping and feeding mechanism inside the mover. This parallel mechanism design effectively shortens the tool length and improves the actuator's anti-vibration capability during polishing.Furthermore, to improve the constant-force output performance, this paper designs a Fuzzy Adaptive Model-Compensated Linear Active Disturbance Rejection Controller (Fuzzy Adaptive Model-Compensated LADRC) to achieve dynamic control of the actuator. Finally, an experimental platform was built, and comparative experiments were conducted to comprehensively test and verify the constant-force output performance of the actuator. The experimental results show that the proposed constant-force actuator exhibits very high stiffness and dynamic response characteristics during polishing operations. Compared to traditional PI control and conventional Linear Active Disturbance Rejection Controllers, it offers faster response speed and smaller adjustment errors. According to the comparison experiments, the proposed control algorithm effectively controls the force output relative error of the mover at different positions within ±1.8%, and the maximum force fluctuation error in polishing surfaces with different roughness levels can be controlled within ±5%, significantly enhancing the automation and quality consistency of the polishing process.