GAIT CONTROL OF HEXAPOD WALKING ROBOTS USING MUTUAL-COUPLED IMMUNE NETWORKS

Biological information processing systems can be said to be one of the ultimate decentralized systems and have been expected to provide vari ous fruitful ideas to engineering fields, especially robotics. Among t hese systems, brain-nervous and genetic systems have already been wide ly used in modeling as neural networks and genetic algorithms, respect ively. On the other hand, the immune system also plays an important ro le in coping with a dynamically changing environment by constructing s elf-non-self recognition networks among different species of antibodie s. This system has many interesting features such as learning, self-or ganizing abilities, etc., viewed from the engineering standpoint. Ther efore, it can be expected to provide novel approaches to the PDP parad igm. However, the immune system has not yet been applied to engineerin g fields. In this paper, we propose a new hypothesis concerning the st ructure of the immune system, called the mutual-coupled immune network s hypothesis, based on recent studies on immunology. We apply this ide a to gait acquisition of a hexapod walking robot as a practical exampl e. Finally, the feasibility of our proposed method is confirmed by sim ulations.