Implementation of Kinematics Bicycle Model for Vehicle Localization using Android Sensors

• Published in: 2020 11th IEEE Control and System Graduate Research Colloquium, ICSGRC 2020 - Proceedings
• Main Author: Ng K.M.; Abdullah S.A.C.; Ahmad A.; Johari J.
• Format: Conference paper
• Language: English
• Published: Institute of Electrical and Electronics Engineers Inc. 2020
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85096945768&doi=10.1109%2fICSGRC49013.2020.9232453&partnerID=40&md5=79601013cce2bf6b986b8c154bcc7881

The kinematics bicycle model is a useful model to perform dead reckoning of vehicles' or robots' dynamic locations. Normally it is implemented using inertial and vehicle odometry sensors. Recently, access to these sensors in the android smart phone could provide a low cost alternative for vehicle localization. Platform such as MATLAB mobile provides functions to access phone sensors. Hence, the objective of this project is to develop vehicle localization by implementation of kinematic bicycle model (KBM) using the MATLAB mobile. Sensors from an android smart phone were accessed via the MATLAB mobile. The kinematics bicycle model is then implemented at two locations, located at Faculty of Engineering, Universiti Teknologi MARA (UiTM) Shah Alam and a residential area in Shah Alam, Malaysia. The position of the vehicle estimated using KBM is continuously logged. The mapped path produced using this logged points will be compared to the actual path to evaluate the performance in terms of mean absolute error (MAE) and root mean square error (RMSE). It was found that the results at the residential area have lower MAE and RMSE due to less multipath effects in the area. © 2020 IEEE.