Experimental research ofprobabilistic localization ofservice robots using range imagedata andindoor GPSsystem

Inthis paper, weproposed aprobabilistic localization of service robot usinglaser scanner andindoor GPSsystem. This scheme isasample based algorithm asanapplication of MonteCarlo Localization. Samples arescattered byreferring toodometry dataandposition dataofindoor GPSsystem. Andthose areevaluated bymap-matching method using laser sensor anddistance error matching using indoor GPSsystem. Weareabletogetmorerobust localization results byusing these processes, eventhough thefloor condition isdeteriorated andthere aremanyorhugeobstacles Anditispossible thatremoveerrors whicharecaused whentherobotis blocked byunexpected obstacles foralong time. Experimental results demonstrate theevaluation oftherobustness ofour algorithm fusing rangeimagesensor andindoor GPSsystem data.