Expression method of human locomotion records for path planning and control of human-symbiotic robot system based on spacial existence probability model of humans

In this paper, a novel describing method of human locomotion trajectory record was proposed, and methods of its application to the path planning and control of human-symbiotic mobile robots were indicated. The novel describing method, named as Existence Record Probability Map (abbreviated ERPM hereafter), was generated through the following three steps: 1) measuring human locomotion path using pressure distribution sensor floor, 2) applying a probability potential model of human body to the measured path, and 3) integrating such probability potential maps for some time period. From an ERPM, a) an existence probability value, b) ridge lines of the potential, and c) Q-value at a point on a ridge line, were elicited. As a usage of these information, methods of a) finding a position where a collision probability of a mobile robot and a human is high, b) finding major trajectories of human locomotion, and c) evaluation of path width to avoid collision with a human, were indicated.