Image source: OFweek.com
Why do robots "shake"?
Why do robots vibrate during operation? We believe the main reasons are as follows:
On the one hand, robot vibration during operation is mainly caused by resonance. On the other hand, the drive control loop is difficult to adapt to all working conditions with a single set of parameters, resulting in a large dynamic range of the load and the inability of the position sensor to directly control the load end, causing end-effector position deviation and vibration. At the same time, due to the presence of flexible components such as reducers , robots are very prone to vibration at the end-effector during positioning.
Why is shaking "difficult" to suppress?
The primary reason is the difficulty in observing vibration. For cost reasons, encoders are generally not installed at the end of the reducer ; the position feedback relies on the output position information of the motor. Secondly, the diversity and time-varying nature of the robot system's on-site operating conditions make vibration difficult to suppress. Finally, factors such as mutual interference of torques between the robot's joints also contribute to the difficulty in suppressing vibration.
7 ways to prevent image stabilization
Through multiple iterations of reasoning, demonstration, experimentation, and feedback, the Peitian Robotics R&D team addressed the robot "shaking" problem with a dialectical approach, developing a series of effective advanced algorithms and successfully overcoming the challenge. Peitian's overall solution incorporates effective methods from various operating conditions.
1. Filter method: Notch filtering is applied to the resonant frequency point, and the phase error caused by the notch is eliminated through gain compensation.
2. Observer method: Predict the position of the end of the body and make some parameter adjustments in advance for the predicted position.
3. Damping method: Extract vibration velocity information and apply a reverse damping to change the vibration response.
4. Driver parameter tuning and dynamic adaptation: Adapt appropriate driver parameters according to the load conditions to achieve effective vibration control.
5. Torque feedforward: Based on dynamics, torque feedforward adapts to load changes with a large dynamic range.
6. Kinematic optimization: Reduce the possibility and degree of resonance by path planning or velocity planning, that is, rationally plan acceleration and jerk for different motion segments to reduce the occurrence of vibration.
7. Reduce excitation intensity: Taking the typical periodic excitation-reducer transmission accuracy change as an example, due to imperfections in processing and installation, the reducer will add periodic interference during transmission, forming a resonant excitation source. By optimizing the quality of core components, it is helpful to control the excitation amplitude and thus suppress vibration.
This solves the problem of robot vibration when moving or stopping, truly improving the quality of industrial robots leaving the factory, providing users with more cost-effective and reliable industrial robot products, and enabling robots to be successfully applied in more industries.
