Sensor fusion Techniques for Automatic Guidance by the method of Kalman filter using DGPS and Gyrocompass (Part 1)

Abstract

We developed sensor fusion techniques combining a gyrocompass and DGPS by the method of Kalman filter for automatic guidance system of an autonomous tractor. Drift or offset of a gyrocompass or heading error is automatically estimated and the accuracy of direction and position of a tractor is improved by the method of Kalman filter, which makes a volume condition of the error of direction and speed data in addition to a position vector using the position data of DGPS as a observation values. We carried out automatic rotary tilling work by the autonomous tractor in the field by using this method. The offset of the optical fiber gyro was correctly estimated and the accuracy of the position estimated was improved over the accuracy of DGPS and able to guide the autonomy along the target line in the field.