Analog based processors can be used to manage all the basic functions involved in the control of an autonomous hexapod walking robot. Generation of locomotion and legs coordination, interaction with the environment (sensors management), execution and control of different tasks can be performed allowing the robot to show biological inspired behaviour. The Cellular Neural Network (CNN) paradigm is particularly suitable to cope with the generation of complex dynamics, thanks to its parallel architecture exploiting the local interaction among simple nonlinear cells. So using CNN it is possible to obtain such kind of behaviour. Starting from this approach it is possible to build modular analog structures with mutual interaction, in order to accomplish each different function in the robot management.
展开▼
