New Trends in the control of Walking Robots

This paper presents some new trends for the control of climbing and walking robots (CLAWAR). In the first part some paradigms concerning the Soft-computing approach for the control of walking robots will be illustrated. Then some considerations concerning the fuzzy control of the robot ROBINSPEC, the self-learning strategy adopted to analyze collective behaviour of the PLIFs microrobots and the fuzzy control of the robot WIRE, are shown. In the third part of the paper the CNN (Cellular Neural Network) paradigm is used to generate and control motion in mechatronic devices. CNNs have been shown to be able to generate active non-linear waves that can be used as the basis to generate and control motion in some devices reflecting architectural paradigm met in animals.