Humanoid robot localization in complex indoor environments

摘要: In this paper, we present a localization method for humanoid robots navigating in arbitrary complex indoor environments using only onboard sensing. Reliable and accurate operating such is challenging task. First, humanoids typically execute motion commands rather inaccurately odometry can be estimated very roughly. Second, the observations of small lightweight sensors most are seriously affected by noise. Third, since walk with swaying freely move environment, e.g., they not forced to on flat ground only, 6D torso pose has estimated. We apply Monte Carlo globally determine track humanoid's 3D world model, which may contain multiple levels connected staircases. To achieve robust while walking climbing stairs, intergrate 2D laser range measurements as well attitude data information from joint encoders. simulated real-word experiments our thoroughly evaluate approach. As illustrate, robot able localize itself accurately its over time.