For biped robot locomotion, two kinds of information are available. One is numerical information from measuring instruments, and the other is linguistic information obtained from biochemical studies of human locomotion. However, each of the above information alone is usually incomplete. In this paper, based on inverse learning, two simple and intuitive integration schemes are proposed to integrate numerical and linguistic information. One is neurofuzzy-based integration, and another is fuzzy rules extraction-based integration. The effectiveness af the proposed methods are verified through biped robot gait synthesis and joint control.
|Title of host publication
|Fuzzy Logic and Intelligent Technologies for Nuclear Science and Industry
|Subtitle of host publication
|Proceedings of the 3rd International FLINS Workshop
|World Scientific Publishing
|Number of pages
|Published - 1998