Heading Estimation for Autonomous Robot Using Dual-Antenna GPS

Abstract

Most of the attitude estimation systems are built from inertial measurement units (IMUs). Micro-electro-mechanical system-based (MEMS) IMUs are low-cost but large errors. MEMS-based angle estimator often uses a tri-axis magnetometer to determine the yaw angle (heading angle) that the estimation accuracy is significantly influenced by the stability of the environment’s magnetic field. This paper introduces a new method to estimate the heading angle using Global Positioning System (GPS) with dual-antenna. The proposed estimation algorithm is independent of the magnetic field and has high accuracy in the heading angle. Through experiments, we also show that the heading accuracy depends on the quality of the GPS receivers and the antennas.