Autonomous Learning by Simple Dynamical Systems with Delayed Feedbacks

TL;DR

This paper proposes a general scheme for autonomous learning in dynamical systems using intrinsic time-delayed feedback, demonstrated with coupled oscillators adjusting their connections to achieve desired synchronization levels.

Contribution

It introduces a novel feedback-based method for dynamical systems to learn and generate specific behaviors without external training.

Findings

The scheme effectively guides systems to desired dynamical states.

Coupled oscillators can adapt their connections to control synchronization levels.

The approach is versatile for different types of dynamical behaviors.

Abstract

A general scheme for construction of dynamical systems able to learn generation of the desired kinds of dynamics through adjustment of their internal structure is proposed. The scheme involves intrinsic time-delayed feedback to steer the dynamics towards the target performance. As an example, a system of coupled phase oscillators, which can by changing the weights of connections between its elements evolve to a dynamical state with the prescribed (low or high) synchronization level, is considered and investigated.