An Extended Kalman Filter for the state estimation of a mobile robot from intermittent measurements

Abstract

Position and orientation estimation is one of the main problem in mobile robotics: to navigate and accomplish its job a mobile robot must know where is it. To this sake, we suppose that a set of wireless sensors are located on the operation field, which can measure distance and orientation of the robot, and are connected with the estimator by a wireless network. An extended Kalman filter has been build up to reconstruct the state of the robot, which into account that the information from the sensors may be lost. Also, in order to save the lifetime of the batteries of the sensors we suppose that the sensors normally sleep (low consumption of energy) and send the information (hight consumption of energy) only if the robot activates them. Therefore, a strategy of scanning of the sensors has been defined, to select at each time which sensors to activate without degrading too much the quality of the estimate.