User directional intention identification for a walking support walker: Adaptation to individual differences with fuzzy learning

Abstract

Safety is required as well as usability when developing a human robot interface for a disabled user. We are developing an omni-directional walker (ODW) to support indoor movement for those who have walking disabilities. A novel method is proposed to recognize a user's directional intention according to his/her forearm pressures to the ODW, which are measured by sensors embedded in the ODW's armrest. Fuzzy rules are extracted from the relationship between forearm pressure and directional intention and an algorithm is proposed for directional intention identification based on distance-type fuzzy reasoning method (DTFRM). Furthermore, fuzzy learning is introduced to adapt to the individual difference in forearm pressures. The experiment results show that the reasoning results of the proposed method are consistent with the intended directions, and that fuzzy learning can reduce the reasoning errors caused by individual difference.