A multiestimation-based scheme for adaptive control of robotic manipulators guaranteeing closed-loop stability

A supervised discrete adaptive control scheme with multiestimation

“…Three parameter adaptive algorithms are proposed with the same relative dead-zone, see [13,15,17,18], in order to show the same results can be reached using different algorithms.…”

“…The following standard assumptions are made in order to then prove the scheme's stability (see [13][14][15][16][17]):…”

“…The main novelty of the proposed approach compared to previous ones in [4][5][6][8][9][10]12,13] is that the various estimation models are of different orders while the corresponding plant unmodeled dynamics is modified accordingly from an estimation model to another one. This permits the design accommodation to the dominant plant frequencies.…”

Intelligent adaptive control of forces in milling processes

“…The strategy is to use the controller parameterization obtained from the best estimation model at each time interval according to the tracking error supervisory action. The closed-loop stability is guaranteed if the time interval between consecutive switchings is larger than an appropriate residence time [8]. The estimated output for each th l identifier can be calculated as, …”

“…The first one tunes each parameter estimation vector through a suitable tuning algorithm (recursive least square, RLS), and, moreover, it can tune an optimal initial β -gain which let to better enhancement transient response behaviors [3]. The second one is related to implement switches through time to online select the correcting gain of the FROH and then the control input plant [8]. At this level, an identification performance index is defined and applied with the aim at finding the most appropriate value of β at each residence time in order to compare the various tracking errors.…”