On the dynamics of small continuous-time recurrent neural networks
Adaptive Behavior - Special issue on computational neuroethology
Distributed Coevolutionary Genetic Algorithms for Multi-Criteria and Multi-Constraint Optimisation
Selected Papers from AISB Workshop on Evolutionary Computing
Incremental evolution of target-following neuro-controllers for flapping-wing animats
SAB'06 Proceedings of the 9th international conference on From Animals to Animats: simulation of Adaptive Behavior
Study on evolution of the artificial flying creature controlled by neuro-evolution
Artificial Life and Robotics
Hi-index | 0.00 |
A new biologically inspired approach to a flapping wing controller which benefits from morphological computation and a Reflexive Pattern Generator (RPG) was tested using a simple physically simulated 3D flying robot. In order to tackle the difficulty of generating robust flapping flight and its manoeuvre, the robot employs simplified flexible "feathers" which are modelled as a series of subpanels attached to the wing skeleton using nonlinear angular springs. The neural controller receives sensory inputs from each feather to let them participate in pattern generation, the robot can also "feel" aerodynamic forces on its wings. From the synergy of flexible feathers and their sensory reflexes, the evolved robot exhibited flight manoeuvre using asymmetric wing movements as well as its tail, and rapidly adapted to external disturbances even in the absence of visual sensors. The reduced stiffness in flight control arising from the wing flexibility is discussed.