User´s moving intention identification for human support robot: Distance-type fuzzy reasoning method based on individual knowldege base

In order to assist lower-limb disabled people to move independently and conveniently, human support robot (HSR) is under development. As a robot directly assist human beings daily, the capability of understanding user´s intention is very important. A novel method is proposed to recognize user´s moving intention according to the information from 4 load sensors embedded in the chair of HSR. Based on experiment, the data features for 9 moving intention were extracted. Combing all the data features, a knowledge base was established for user´s intention. For distance-type fuzzy reasoning method (DTFRM), rules were built on the knowledge base which was based on the statistical property of information from load sensors. The DTFRM was applied in this paper to reason user´s moving intention. Finally, control experiments were performed to verify the identification performance of this method. The experiment results show that this method can identify user´s moving intention based on his action habits. Then HSR can carry user to move in the intended direction exactly without any control by hand. The distance-type fuzzy reasoning method is suitable to infer user´s intention from information of load sensors caused by user´s action.