Controlling an inverted pendulum by neural plant identification

The paper shows how well-known supervised learning techniques can be applied to learning of unstable systems control. This is presently done in two steps. In a first step, a neural network is trained to predict the dynamic behavior of an arbitrary plant as accurate as possible. In a second step, the neural model is expanded by a control part in order to learn a control strategy that leads the system's state variables on given trajectories.<>