Gene regulatory network to control and simulate virtual creature's locomotion

Abstract

Abstract: - This paper explores the application of evolutionary computation technics for evolving behaviours of virtual creatures inhabiting a realistic virtual environment. Our approach uses a computational model of gene regulatory network, which is inspired of cell control mechanism of real cells. Usually used to control virtual cells in developmental models, recent works showed that gene regulatory networks are also capable to control various kinds of agents. This paper details how a gene regulatory network is evolved to control range of articulated virtual creatures. To do so, the inputs and outputs of the network are directly mapped to the creatures’ sensors and actuators. To evaluate this approach, we have compared its performance to one of the most recent successful evolutionary method, the NEAT algorithm. The results show the gene regulatory network model may possibly be a viable solution for evolving control solutions for physical machines.