Localization based on a 3D extended space observation model


This paper describes a localization technique for a mobile robot in an environment with many unknown obstacles, such as pedestrians. To realize robust localization against unknown obstacles by using a particle filter, a space observation model has conventionally been used. Although this conventional model is very effective, it only uses two-dimensional information, which is dependent on the scan plane of laser scanner. If a model is extended to three dimensions, the observed information increases, and the the accuracy of the position estimation improves in comparison with that of the two-dimensional model. In this research, the proposed model extends the conventional model to three dimensions. Therefore, the proposed model still has the advantages of the previous model; namely, the likelihood of a particle can be evaluated by a simple method using a particle filter, and the occupied space can be considered. In a real environment, it was shown that the proposed model is more effective than the two-dimensional conventional model.

2012 JSME Conference on Robotics and Mechatoronics (ROBOMEC2012)