Generation of Avoidance Motion for Robots which Coexist and Cooperate with Humans. 2nd Report. Avoidance of Moving Obstacles for Mobile Robots.

Abstract

This paper presents a trajectory planning method by which human cooperative mobile robots can avoid collision with human bodies. Indefinite human motion is predicted using a stochastic predictor. Evaluation of collision risk and efficiency of trajectory are taken in account in an evaluating function. Optimization using a genetic algorithm gives the robot trajectory. The mobile robots move so as to maximize safety taking into account every possible human motion. Simulation results showed that the robot could determine appropriate trajectories avoiding danger and that reasonable trajectories were generated even if human motion unexpectedly changed.