Improving Path Accuracy and Vibration Character of Industrial Robot Arms with Iterative Learning Control Method

Iterative learning control (ILC) enhances control specifications for display panel transfer robots used continuously in industrial settings, offering significant cost-effective improvements in sites requiring higher control standards. This study focuses on improving the path accuracy of an 8G display panel transfer robot control system, as defined in ISO 9283, by implementing a path error compensation ILC system. To mitigate the time-delay characteristic of path error compensation inputs in the robot system, an ILC algorithm was developed. It incorporates time-scaling and time-shifting algorithms in addition to the switching-gain, proportional, derivative (SPD) offline ILC method. Furthermore, a compensation system was designed to ensure the stability of the compensation input, integrating a low-pass filter into the proposed ILC algorithm. Experimental validation of the ILC compensation system was conducted using an 8G display panel transfer robot, demonstrating its functionality. Additionally, the ILC path error compensation system parameters were optimized through various experiments and detailed characteristic analyses. Iterative learning resulted in a reduction of RMS path error data by more than 90%, significantly enhancing control performance.