On control structures for position controlled robots using time-optimal and model-referenced adoptive control concepts