Auto-tuning of lead step in phase lead iterative learning control

In phase lead compensation Iterative learning control, lead step provides a phase lead to compensate the phase lag caused by feedback controller. With this phase lead compensation, the monotonic decay of tracking error can be guaranteed in a widened frequency band. Then, the learning performance can be greatly improved. Because of the inaccuracy of system model, the optimal lead step often cannot be obtained. In this paper, the tuning of lead step based on learning transient behavior is introduced such that the optimal lead step can be obtained automatically in the case the system model is not available. With this tuning scheme, the dependence on system model can be eliminated. To verify the tuning scheme, it is carried out on an industrial robot and the experimental results show that the tracking error during this tuning process keeps at a very low level.