18 June 2019 in IEEE Transactions on Neural Networks and Learning Systems
In the fields of artificial neural networks and robotics, complicated, often high-dimensional systems can be designed using evolutionary/other algorithms to successfully solve very complex tasks. However, dynamical analysis of the underlying controller can often be near impossible, due to the high dimension and nonlinearities in the system. In this paper, we propose a more restricted form of controller, such that the underlying dynamical systems are forced to contain a dynamical object called a heteroclinic network. Systems containing heteroclinic networks share some properties with finite-state machines (FSMs) but are not discrete: both space and time are still described with continuous variables. Thus, we suggest that the heteroclinic networks can provide a hybrid between continuous and discrete systems. We investigate this innovated architecture in a minimal categorical perception task. The similarity of the controller to an FSM allows us to describe some of the system's behaviors as transition between states. However, other, essential behavior involves subtle ongoing interaction between the controller and the environment that eludes description at this level.
30 October 2022 in Dynamical Systems
A heteroclinic cycle is an invariant set in a dynamical system consisting of saddle-type equilibria and heteroclinic connections between them. It is known that deterministic perturbations (inputs) to a heteroclinic cycle generally lead to periodic solutions. Addition of noise to such a system leads to a non-intuitive result: there is a range of noise levels for which the mean residence time near the equilibria of the heteroclinic cycle increases as the noise level increases to a given threshold. We explain how the interaction between noise and inputs gives rise to this by combining analytical results from constructing a Poincaré map with a simple stochastic system. We support our results with numerical simulations.