Research about an estimated position of a mobile robot used Plug and Play estimating position system

摘要

In this paper, we proposed a estimating position method using multi-sensor for a mobile robot. This estimating method takes on a sensor fusion using Kalman filter and decentralized estimating method. In the latter method, each sensor considered sensor unit that has arithmetic capacity for estimating self-location. Each sensor unit estimates own position and orientation. And this estimated position and attitude are exchanged with other unit, and done sensor fusion by Covariance Intersection. By doing so, it's able to build the system that can estimate self-location correctly if an user changes a sensor.