High-order natural iterative learning control of robotic manipulators: new results

Abstract

In this paper it is considered problem of control robotic systems using suitable iterative learning control. Recently, there have been extensive research activities in the topic of learning control for controlling dynamics non-linear systems in an iterative manner. The learning control concept differs from conventional control methodologies in that the control input can be appropriately adjusted to improve its future performance by learning from the past experimental information as the operation is repeated. Iterative learning control (ILC) requires less a priori knowledge about the controlled system in the controller design phase and also less computational effort than many other kinds of control. Dynamic model of robotic manipulator with uncertainties can be presented in the form of state space and output equations as a class of time-varying, non -linear system. Motivated by human learning, the basic idea of iterative learning control is to use information from previous executions of the task in order to improve performance from trial to trial in the sense that the tracking error ei (t) is sequentially reduced. The learning control scheme comprises two types of control laws: feedback law and feed-forward control law. It is proposed high-order feedforward ILC using local negative feedback on control with great amplifying. Sufficient conditions for the convergence of a new type ILC algorithm for a class of time-varying, non-linear system are obtained. Finally, results are applied to a suitable robotic manipulator through simulation, which demonstrates the effectiveness of the method.