To solve the problem that using the information of the robot vision or sonar system can not give good self-localization result

a new self-localization method for autonomous mobile robot(AMR) is proposed under polar coordinates.In this way

the distances and the angles of vision

driving motors and the sonar respectively can be derived from the simple geometrical algorithm

then

using the method of fuzzy mathematics to fuse the information.Thus the ideal distance and the angle information can be obtained to improve the self-localization accuracy of AMR.Experimental results show that the localization error is lower than 1.5% in the range of 4 m

and that the method provides a reference for the self-localization of multiple robots.