Experimental 2D extended Kalman filter sensor fusion for low-cost GNSS/IMU/Odometers precise positioning system

The development of satellite techniques and the availability of mobile devices with built-in multi-GNSS (Global Navigation Satellite System) receivers allow the determination of position with increasing accuracy. At the same time, the requirements of users as to the accuracy of positioning are increasing, while the low production costs of the device must be maintained. This paper presents a loosely coupled integration of low-cost sensors (GNSS, IMU (Inertial Measurement Unit), and an odometer) with the use of a nonlinear Kalman filter and a dynamic weight matrix. The integration model was developed for horizontal (2D) components with the simultaneous determination of the azimuth of the test platform. The tests were carried out in the conditions of an open horizon, with partial obscuring of the horizon (passage under an open-work steel structure) and along walls. In this way, the working conditions for an autonomous lawn mower, which are now increasingly used by citizens, were simulated. The position accuracy obtained in these tests is better than 5 cm for horizontal components and better than 1 degrees for the azimuth.