Abstract
Problem of stabilizing unstable equilibria was discussed using discrete controls. Discrete control created a chaotic attractor in the vicinity of equilibrium. Artificial neural networks with reinforcement learning were known to be able to learn control schemes.