Controlling Unknown Saddle Type Steady States of Dynamical Systems with Latency in the Feedback Loop

We suggest an adaptive control technique for stabilizing saddle type unstable steady states of dynamical systems. The controller is composed of an unstable and a stable high-pass filters operating in parallel. The mathematical model is considered analytically and numerically. The conjoint controller is sufficiently robust to time latencies in the feedback loop.

In addition, it is not sensitive to the damping parameters of the system and is relatively fast. Experiments have been performed using a simplified version of the electronic Young-Silva circuit imitating behavior of the Duffing-Holmes double-well oscillator.