A Model Chemical Memory in an Evolved Animat

This paper describes work carried out to investigate whether a classic simulated reaction-diffusion (RD) system could be used to control a ‘minimally cognitive’ animat during the course of a simple memory test. This test required the animat to remember an arbitrary signal and adapt its behaviour as a result. A further requirement was that the effects of the first signal be reversed by a second signal, returning the animat to its default behaviour. In this way the two signals combined to form a behavioural-switch, regulated by a memory-trace preserved in the homogeneous chemical substrate. The reactiondiffusion system chosen was that first described by Gray and Scott (Gray-Scott) and the minimally cognitive behavior of a class introduced by Beer et. al, involving the fixation and avoidance of a falling circular object by a whiskered animat. The parameters of this RD-controller were evolved using an evolutionary, or genetic, algorithm (GA).