Abstract :
To increase intelligence of walking-assisted devices, it is important to provide a start command of action based on the motion intention of subjects. In this paper, a method for identifying motion intention was proposed based on cerebral hemoglobin information. Spontaneous movements (upstairs, downstairs, upslope, down-slope, sit-down, squat, and the corresponding standup) without an advanced imagination or external stimuli were performed on two subjects. During the experiment, cerebral hemoglobin information was recorded by applying near-infrared spectroscopic technology. Multiple analyses of variance on different features of the variation of total hemoglobin were combined and a series of significant levels were set. The corresponding precision rate was up to 74.8%, and the sensitivity was up to 79.2%. Since the proposed identification method is realized based on the brain information recorded in a real movement environment, it is helpful and practical to enhance intelligence of walking-assisted devices.
Keywords :
"Legged locomotion","Detectors","Acceleration","Analysis of variance","Probes"