Stage positioning based on improved feedback error learning and its experimental verification

In this paper, the position control of a linear stage is considered. To improve the settling time of the positioning, the parameters of the feedforward controller are automatically tuned by using feedback error learning (FEL). In the proposed FEL, true value of the estimated parameters is small, and the number of the estimated parameters is reduced. It leads to the improvement of the learning time. Moreover, to compensate for effects of non-parallel guide rails, the disturbance observer is also utilised. By using the observer, the performance of the positioning can be maintained regardless of the initial position of the stage. Experimental results show the effectiveness of our control method.