A Low-Cost Positioning System for Parallel Tracking Applications of Agricultural Vehicles by Using Kalman Filter

A position-velocity (PV) model and a multi-sensor system, consisted of a consumer application GPS, a MEMS gyro, two encoders, and a turning angle sensor, was constructed for the positioning system. The two encoders augmented the positioning accuracy greatly that the fluctuation of vehicle position was greatly smoothed comparing with a GPS-only system. The minimal fluctuation was falling from 2.21 m to 0.52 m (east direction), from 0.68 m to 0.23 m (north direction). The maximum XTE was reduced from 2.5 m to 0.77 m, and the RMS value was improved to 0.22m. The GPS bias error was the major difficulty to produce better performance.