Abstract

The desirable autonomy in a mobile robot, force to identify its position and orientation "pose", in relation to a global reference system previously defined. Without "a prion" assumptions, i.e. a region for start of search, the problem of pose acquisition has a combinatorial feasibility potential that can represent a serious challenge to naive attempts of solution. Even in the simple case of knowledge of the most probable starting region, myopic procedures inspired in steepest descent do not succeed in many trials. This paper presents a procedure of pose acquisition based on Evolutive Strategies and afterwards maintaining the pose using a myopic procedure here presented as a result of potential minimisation between measurement points and a CAD map segments. The switch from pose acquisition to pose correction is made depending on the closeness measure obtained at each step. If the closeness value exceeds a given bias, the procedure of pose acquisition follows and thereafter the pose correction takes place.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call