Implementation of Iterative Learning Control Method for Ultrasonic Intermittent Welding

This study proposes a robust iterative learning control (RILC) methodology to address the demand for rapid and stable amplitude regulation in intermittent wave generation processes. The proposed approach synergistically integrates iterative learning control principles with robust control theory, enhancing control precision and dynamic response through systematic iterative parameter optimization. Furthermore, the design incorporates robustness-enhancing mechanisms to counteract practical uncertainties including detuning effects and external pressure variations inherent in real-world welding operations, thereby ensuring system stability and control reliability under dynamic pressure conditions. Experimental simulations demonstrate the superior performance of the RILC scheme compared with the extended state observer (ESO) approach under maximum detuning conditions. Notably, the proposed scheme restricts maximum overshoot to below 0.5% while achieving a 4ms rise time – representing a 50% improvement in response speed compared to conventional ESO-based methods.