A Bias Compensation Strategy for Wheeled Mobile Robot Odometric Self-localization Algorithm

In service robot applications, the automated in-door mobile robot should be able to localize itself by the equipped sensor measurements. This paper is concerned with the wheeled mobile robot self-localization using the digital compass and photoelectrical encoder measurements. A traditional odometric localization algorithm is utilized and the bias analysis of this method is derived based on the statistic theory. Then, a low cost bias compensation strategy is developed to improve the estimation performance. Also, a classical extended Kalman filter (EKF) is applied as a comparison. The effectiveness of the proposed self-localization algorithm is verified with simulation examples.