Digital filters to integrate global positioning system and dead reckoning

The design, test and results of a positioning system consisting of a differential global positioning system (DGPS) receiver, a radar velocity sensor, a wheel velocity sensor and an electronic compass are discussed. The characteristics of the individual sensors were investigated and with these results a Kalman filter was designed to integrate the different signals into a single estimate of position, velocity and heading. Field tests were conducted to investigate both the real-time performance of the system and the robustness against distortions caused by malfunctioning of the sensors. An extended Kalman filter was found to improve accuracy of positioning and was able to bridge gaps in position information due to blockage of the GPS receiver. The position measurements from the GPS receiver had a standard deviation of 1.18 m. The use of a nine-state filter reduced the standard deviation of positioning to 0.71 m. On a track which ran along a line of trees and between buildings, the filter was able to maintain position for 12 s, while the GPS receiver lost satellite fix. (C) 1998 Silsoe Research Institute.