Pneumatic Compliant End-effector Using Multi-compensation Integrated Nonlinear Active Disturbance Rejection Control
ObjectiveIn robotic polishing operations, an end-effector polishing mechanism is typically used to achieve constant force control. Therefore, the performance of this control directly affects the quality and efficiency of the polishing process. Among various end-effector mechanisms, the pneumatic typ...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Editorial Department of Journal of Sichuan University (Engineering Science Edition)
2025-05-01
|
| Series: | 工程科学与技术 |
| Subjects: | |
| Online Access: | http://jsuese.scu.edu.cn/thesisDetails#10.12454/j.jsuese.202300718 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | ObjectiveIn robotic polishing operations, an end-effector polishing mechanism is typically used to achieve constant force control. Therefore, the performance of this control directly affects the quality and efficiency of the polishing process. Among various end-effector mechanisms, the pneumatic type is widely adopted due to its lightweight structure, low cost, robustness, and ease of maintenance. This paper focuses on achieving constant force control in a pneumatic end-effector compliant mechanism.MethodsAchieving constant force control in this mechanism is challenging due to nonlinear factors within the pneumatic system (such as gas compression, proportional valve dead zones, and cylinder friction) and the effects of posture changes on output force. A novel control algorithm is proposed, integrating five components: nonlinear active disturbance rejection control, gravity compensation, dead zone compensation, LuGre friction model compensation, and data filtering.Nonlinear active disturbance rejection control: A tracking differentiator preprocesses the input signal to reduce overshoot, yielding a tracking signal and its derivative. An extended state observer estimates total system disturbance in real time. A nonlinear state error feedback law performs error correction and compensates for the observed disturbance. Gravity compensation: A gravity compensation strategy is developed based on the relationship between gravitational force on the mechanism and its output force at different polishing angles. Dead zone compensation: Minimum and maximum working voltages of the proportional valve are experimentally determined. These voltages are used to design mid-position compensation, reducing the dead zone’s impact on control performance. LuGre friction model compensation: Friction data is collected under varying cylinder speeds. LuGre model parameters are fitted, and a compensation strategy is developed to offset frictional disturbances during piston movement.Data filtering: A first-order low-pass filter is integrated into the controller to reduce high-frequency noise in the force sensor readings and enhance control accuracy.The control algorithm is implemented on an STM32F103 microcontroller operating at 50 Hz. Supporting circuits for the pneumatic actuator are designed, and a test bench is constructed to simulate robot polishing conditions. To mimic posture changes during polishing, the platform angle is manually adjusted using a screw. In addition, to simulate the position errors that may occur during the robot’s motion trajectory planning, a stepper motor with a ball screw is used to control the movement of the connection end of the pneumatic end-effector compliant polishing mechanism closer to or further away from the polishing surface. A variety of working condition simulation experiments were conducted, including:1) constant force loading experiments and sinusoidal force loading experiments with the pneumatic end-effector compliant mechanism in a vertically downward position; 2) variable-angle constant force loading experiments involving changes in the polishing posture of the pneumatic end-effector compliant mechanism; 3) external disturbance loading experiments involving vertical movement of the connecting section of the pneumatic end-effector compliant mechanism.Results and DiscussionsUnder a set force of 50 N, the designed controller exhibited an average error of 0.21 N and a standard deviation of 0.18 N. This performance is better than that of PID control, which had an average error of 0.27 N and a standard deviation of 0.21 N. In two sinusoidal force tracking experiments with periods of 8 s and 4 s, the results showed that for the 8 s period, the designed controller’s tracking performance was comparable to PID control, but with smaller tracking error at extreme points. For the 4 s period, the PID-controlled tracking curve exhibited distortion, whereas the designed controller showed some lag but remained closer to the desired values. Under a 50 N loading with polishing angles varying from 0° to 75°, the controller with gravity compensation had a maximum error of 1.44 N. In contrast, the controller without gravity compensation showed a significant decline in end-effector output force as the polishing angle increased, with a maximum error of 8.82 N. Under the same loading with disturbances present, the designed controller achieved a maximum error of 7.24 N, compared to 11.79 N with conventional disturbance rejection control and 14.77 N with PID control. These results indicate that the designed controller performs better in the presence of disturbances.ConclusionsThe experimental results demonstrate that the proposed control algorithm offers superior robustness, tracking performance, and disturbance rejection compared to traditional PID control. In addition, it effectively compensates for changes in polishing angle, thereby improving the constant force control performance of the pneumatic end-effector actuator. |
|---|---|
| ISSN: | 2096-3246 |