Skill-assist Control of an Omni-directional Wheelchair by Neuro-fuzzy Systems Using Attendants’ Force Input

This paper presents a novel skill-assist (or power-assist) method which employs neuro-fuzzy systems in an Omni-directional Wheelchair (OMW). This function allows the attendants of elderly or handicapped people to deal with heavy loads while the operability problem of moving the OMW laterally or rotating it around the OMW’s Center of Gravity (CG) is also solved. A fuzzy reasoning method is proposed for estimating the navigation direction according to the force exerted by the attendants against the handgrips of the OMW. A neuro-fuzzy system (ANFIS) is used for auto-tuning the fuzzy system’s membership functions according to each attendant’s individual characteristics based on the attendant’s input data. Further, a second-order lag controller which transforms the force applied by the attendant into the OMW’s velocity is presented for enhancing the comfort of the OMW’s occupant. The usefulness of the proposed method is demonstrated through simulation and experiments.