Predictor-Network-Based MRACS for Inverted Pendulum System
暂无分享,去创建一个
There have been many approaches in the field of stabilizing control of inverted pendulum system. In many cases, however, linealization technique for small deviation from equilibrium point has been adopted to apply LQ theory or MRACS to the inverted pendulum system. These algorithms are not necessarily satisfactory for a class of non-linear and time-varying systems such as the inverted pendulum system.In this paper, we propose a new MRACS technique with predictor network which is applicable for the non-linear and time-varying system, inverted pendulum system. This predictor-network-based MRACS can also achieve any convergence characteristics.
[1] R. Monopoli. Model reference adaptive control with an augmented error signal , 1974 .
[2] C.W. Anderson,et al. Learning to control an inverted pendulum using neural networks , 1989, IEEE Control Systems Magazine.