Stable Adaptive Control with Recurrent Neural Networks
| Authors: | Lefebvre Dimitri, University Le Havre, France
Zerkaoui Salem, University Le Havre, France
Druaux Fabrice, University Le Havre, France
Leclercq Edouard, University Le Havre, France |
|---|
| Topic: | 2.3 Non-Linear Control Systems |
|---|
| Session: | Nonlinear Controller Design |
|---|
| Keywords: | Adaptive control, fully connected recurrent neural networks, Lyapunov function, multivariable systems, stability analysis. |
|---|
Abstract
In this paper, stable indirect adaptive control with recurrent neural networks is presented for multi-input multi-output (MIMO) square non linear plants with unknown dynamics. The control scheme is made of a neural model and a neural controller based on fully connected RTRL networks. On-line weights updating law, closed loop performance, and boundedness of the neural network weights are derived from the Lyapunov approach. Sufficient conditions for stability are obtained according to the adaptive learning rate parameter.
