Online Calibration Method of Smartphone Magnetometer in Vehicle Geomagnetic Matching Positioning

Abstract

Geomagnetic navigation has been a potential technology for the indoor positioning of smartphone users, especially in the indoor parking area. The geomagnetic matching positioning process can be divided into two parts: offline geomagnetic map construction and online geomagnetic matching, which both require the geomagnetic vector data obtained to match in the same coordinate system without magnetic interference. Compared with geomagnetic positioning for pedestrians, when smartphones conduct magnetic vector measurement in vehicles, there are problems such as uncertain placement posture and vehicle's own magnetic field interference, and it is difficult to implement conventional magnetic measurement calibration methods. Therefore, the online calibration method of smartphone magnetometer is proposed for vehicle geomagnetic matching and positioning. Firstly, the vehicle motion mode recognition and quasi-static magnetic field detection are applied for data selection and division. Then, the data of stationary mode is used to estimate the horizontal misalignment, and the data of straight-moving mode is used to estimate the heading misalignment and magnetic interference. The trust region algorithm is used for parameters calculation, which produces desired results with little computation. Experiment shows that the measurement difference of X-axis and Y-axis decreases to 0.2476uT and 0.0153uT when the smartphone's position and posture are different.