Autonomous navigation for an unmanned mobile robot in urban areas

Abstract

Mobile robot has attracted intense research interest in recent years due to numerous potential application. The capability of autonomous navigation is critical for a mobile robot when performing a wide range of military and civilian tasks in urban areas. An autonomous navigation system for an unmanned mobile robot in urban areas is presented in this paper, including location subsystem and navigation tracking subsystem. An inexpensive single GPS receiver and inertial sensors as gyro, compass and encoders are deployed in the location subsystem. The important contribution is the GPS/DR fusing algorithm, which tries to fuse the information coming from GPS and DR. Meanwhile the kalman filters are introduced to smooth the GPS location information based on the Markov model and improve the accuracy of Dead Reckoning(DR) location. In the navigation tracking subsystem, the heading angle tracking is achieved by subdivision speed control and trajectory linearization control algorithm is designed to follow line path. Then the errors of east, north direction and distance, are put forward to describe the performance of the robot location and navigation. The experiments carried out on the differential drive tracked mobile robot demonstrate that the whole navigation system is feasible and reliable.