Bio-Inspired Emergent Control of Locomotion Systems

Even if animals deprived of commands from high-level neural centres are still able to locomote, in many cases posture control is lost. This is clear evidence of the fact that posture control is a complex form of behavior involving high-level control and feedback from the visual system, vestibular system, and so on. In the motor system the role played by high-level control centers, also called command neurons, can be very complex. In this and in the next Chapter we will give some examples of bio-inspired high-level control. In particular, this Chapter deals with the important issue of attitude control for the hexapod robot approached with two different strategies. The first one is based on distributed control through CNNs; the second, based on Motor Maps, introduces a more general framework for the control of nonlinear systems in which self-organization and unsupervised learning play a fundamental role in realizing an adaptive controller inspired by the motor cortex of the human brain.
This Section focuses on a biologically inspired analog control system to solve the task of attitude stabilization and locomotion control in a hexapod robot. Attitude control of a rigid body in three-dimensional space consists of leading the object to a desired orientation with respect to a fixed reference frame by computing a suitable control law (for a survey on the attitude control problem see [Kreutz-Delgado and Wen (1991)]). Of course this issue is fundamental in hexapod robots when walking on uneven terrain.
In [Uchida et. al.