A study on precognition of walking patterns for a power assist robot legs

Abstract

In recent times, The assist robots are widely used to assist walking of the old and the weak and disabled as well as to rehabilitate the injured by helping with muscle strengthening exercises. In this study, we present walking patterns method of precognition for the control of the walking-assistance robot legs for the old and the infirm. We extracted the precognition signals of walking patterns using the union of EMG(Electromyogram)signal and EEG(electroencephalogram) Signal for smooth operation of robots by judging the operation patterns in advance when people and equipped with the Assistance robot. We also proposed HMM(Hidden Markov Model) parameter of re-estimation using Forward-Backward Procedure Algorithm. We believe that this signal can be used for the effective operation of Assistance robots.