TABLE OF CONTENTS The neurons belonging to the CPG network have been referred to in this Chapter as motor-neurons. The reason for this is that they are directly used to drive the legs of the hexapod robot. Many biological models also include inter-neurons: these neurons are involved in the generation of the rhythm of the locomotion and are connected to the motor-neurons that directly synapse the muscles. The approach presented here is easily generalized to include this scheme: it can be assumed that the neurons of the CPG simply give the rhythm for the locomotion system, while other motor-neurons are included in a hierarchical structure, receiving signals from the CPG and driving the actuators. The approach depicted is therefore suitable for the control of a more general locomotion system, with a given number of n actuators, driven by n motor-neurons of a CPG with N inter-neurons. Moreover, the assumption that the neurons of the network are identical can be removed.
Therefore, two generalizations leading to a more flexible CPG design can be introduced:
including inter-neurons;
considering neurons with different parameters.
An example of such a CPG is illustrated below. The half-bound of the hare [Collins and Stewart (1993a)] is a locomotion pattern in which the back legs move synchronously and the front legs are half a period out of phase with the back pair, and slightly out of phase with one another. This pattern is adopted by some animals at galloping speeds (for instance, squirrels). To...
