원문정보
초록
영어
The repeatability of system is a fundamental requirement for various iterative learning control methods, and is a necessary condition for the outcome of perfect tracking. This paper theoretically and numerically explains that how the history before the initial time of dynamic systems influences the current state and repeatability of the system. To this end, the convergence analysis of PD-type iterative learning control for initialized system is presented. A practical preconditioning strategy is added to accelerate the convergence speed, and the detailed discussions of initialization function and initialization response are shown as well. The minimum preconditioning time interval is achieved, and some unique properties of initialized system are illustrated to provide novel challenges for robust and adaptive controls. A number of numerical simulations exhibit that a simple preconditioning process can efficiently improve the performance of the initialized iterative learning control.
목차
1. Introduction
2. Preliminaries
2.1. Laplace Transform
2.2. Initialized System and Preconditioning
3. The Initialized MIMO System
3.1. Zero Initialization Function
3.2. Finite Time Initialization Function
3.3. Infinite Time Initialization Function
3.4. Initialization Response
4. Convergence Analysis
4.1. The Case of Zero Initialization Function
4.2. The Case of Non-Zero Initialization Function
5. Illustrated Examples
6. Conclusions
References