返回
Iterative learning neural network control for nonlinear system trajectory tracking
DOI:10.1016/S0925-2312(01)00661-0.png)
摘要
En 中文
This paper presents a neural network control scheme for tracking a non-periodic trajectory with a finite time interval from the view point of the iterative learning control. A system with time-varying but iteration-invariant uncertainties can be controlled by this scheme. The controller consists of a series of local networks. Every point along the desired trajectory has its own one for approximating the nonlinearity in a neighborhood only. An iterative training law described by a positive definite discrete kernel is presented. The control scheme is applied to robot movement imitation by visual servo. Stability of the controller is ensured. (C) 2002 Elsevier Science B.V. All rights reserved.
Keyword:
iterative learning control
neural network control
adaptive control
visual servo

