Data-Driven Model-Free Iterative Tuning Approach for Smooth and Accurate Tracking

The feedback controller tuning for tray indexing can be challenging and time-consuming depending on the experience of the control engineer. Inappropriate tuning of the feedback controller often results in unacceptable tracking accuracy, humming sound, and even chips dislodged. In this paper, we propose a data-driven iterative feedback tuning approach within a three-degree-of-freedom control structure. The penalty function is introduced which takes into account both the tracking accuracy and jerk minimization, and it is particularly suitable for the tray indexing application. In contrast to the traditional model-based control, the proposed approach method makes use of the closed-loop experimental data to fine-tune the controller without system modeling, i.e. a model-less approach. The performance improvement is demonstrated with a case study based on a typical industrial timing-belt setup.