Sensors fault diagnosis in autonomous mobile robots using observer — Based technique

Abstract

This research investigates a sensors fault diagnosis in autonomous mobile robot. Through this study we use the observer - Kalman filter identification technique. According to this observer - based method, the experimental input-output data are being exploited for system identification. The motive is to design observers exclusively for each sensor of the system that allows generating residuals. The research goal is to provide early sensors fault diagnosis. In order to prove the efficacy of the proposed method, we investigated different type of faults and an experimental procedure was carried out using a Pioneer 3AT mobile robot.