An Autonomous vehicle is a driverless vehicle which is an emerging area in automotive industry across world. In an autonomous vehicle, the Advanced Driver-Assistance System (ADAS) is an indispensable component which is developed to automate, adapt and enhance vehicle systems for safety and better driving. The key factors in the ADAS are reliable vehicle's position and orientation. In the present system, Global Positioning System (GPS) receivers with a real time kinematics (RTK) are used in autonomous vehicle. But the reliability of GPS signal and orientation accuracy is not satisfactory because of its inherent limitations. In this paper work, the Inertial Measurement Unit (IMU) which is combination of sensors and GPS are integrated with Extended Kalman Filter (EKF) to overcome the limitations in the present system, thus a navigation algorithm is proposed for autonomous vehicles. The proposed algorithm methodology involves recording of the data from the GPS and IMU simultaneously and processing those obtained data in the EKF to reduce the accumulated errors. Then it is integrated in the iterative process of the filter, which gives the final positional data. In the driver less environment, the Dead Reckoning (DR) is a condition when there is a loss of GPS signal and the vehicle position depends on IMU data. The on-field performance trials with DR are conducted and evaluated for different real-time urban scenarios with both simulated results and pre recorded data on the demo path.
Vehicle state estimator for urban autonomous solution POC
01.08.2017
1391903 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
Central State Estimator for Vehicle Control Systems
British Library Conference Proceedings | 2007
|