In this paper, we propose a new concept-direct learning which is defin
ed as the generation of the desired control input profile directly fro
m existing control input profiles without any repeated learning. The m
otivation of developing direct learning control schemes is to overcome
the limitation of conventional learning control methods which require
that the desired tracking patterns (trajectories) be strictly repeata
ble throughout the learning process. The main advantages of the direct
learning are (1) the capability of fully utilizing the pre-obtained c
ontrol input signals which may correspond to tracking patterns with di
fferent magnitude scales and be achieved through various control appro
aches; (2) direct generation of the desired control input profile with
out repeating the operation cycles. The focus of this paper is on dire
ct learning for a class of trajectories which have identical operation
periods but are different in magnitude scales.