Cooperative Localization Using Fuzzy Decentralized Extended Information Filtering for Homogenous Omnidirectional Mobile Multi-robot System

Abstract

This paper presents a cooperative map-based localization method using fuzzy decentralized extended information filtering (FDEIF) for a multi-robot system composed of multiple homogeneous omnidirectional mobile robots (OMRs). With the help of one KINECT-based landmark or QR code found by one robot, a pose initialization algorithm is applied to estimate both start-up positions and orientations of that robot, and a cooperative localization method is proposed to find initial poses of other robots that can be directly detected by a laser scanner mounted at this localized robot. Once the initial global poses of all robots have been roughly determined, an FDEIF dynamic cooperative localization method is proposed to fuse external measurements from one RGB-D sensor, or one laser scanner and four encoders mounted on each robot, in order to continuously keep track of the accurate moving poses of all the robots. Simulations are conducted to show that the proposed method is effective in finding accurate estimation of both unknown initial and continuous moving poses of three omnidirectional mobile robots in a team.