Implementasi Sensor Inertial Measurement Unit untuk Sistem Odometri Kendaraan Otonom

Abstract

Odometry enables a moving object to be estimated its position. Odometry is the determination of the movement of a vehicle based on sensor data taken periodically. Positioning is needed in decision making in autonomous vehicles. Accurate positioning can result in making the right decisions in autonomous vehicles, resulting in safe operations on the traffic. Odometry on vehicles generally uses wheels that are connected to an encoder. Through this system, errors in determining position are accumulated due to loss of traction between the wheels and the road. An odometric system that does not require contact with the road is required to reduce mechanical losses due to slip. In this study, an odometric system is proposed that measures vehicle displacement based on the vehicle acceleration measured periodically. To measure vehicle acceleration, the odometry system is equipped with an IMU sensor. ARM-based microcontroller is used to acquire acceleration data and perform displacement calculations. Vehicle displacements are estimated using Kalman Filter. This research discusses the performance of using the IMU sensor for the odometry system through experiments. Based on the tests that have been carried out, the resulting accuracy of the proposed odometry system reaches 63.7%.