A Comparison of Information Fusion Methods for Locating Intelligent Mobile Robot

Self-localization methods for intelligent mobile robot can be found from literatures. Here, we studied two information fusion methods, namely extended Kalman filter and unscented Kalman filter. They are used to locate the pose of mobile robot that is navigating in the indoor environment. To analyze the performance of the two filters, they were used respectively to fuse the information coming from the onboard odometry and unidirectional camera. We built the nonlinear models for these two sensors and studied the propagation of uncertainty transformed by the given nonlinear system. Finally we drew a comparison between the two approaches based on the SmartROB2 mobile robot and the performance analyses are given accordingly