The Research Of Data Fusion Based On GPS/DR Integrated Positioning System

Vehicle location technology is the core part of the intelligent transportation system,and a combination of navigation and positioning is the main trend of vehicle locationtechnology. GPS(Global Position System,GPS) with all-weather, real-time, highpositioning accuracy, but the signal is likely to be obscured. Dead Reckoning (DeadReckoning, DR) system is a self-positioning technology, do not rely on external signals,but the error is easy to accumulate, will lose positioning effect for too long. ThereforeGPS/DR integrated positioning system will combine GPS and DR from each other, sothat the positioning accuracy of combined is higher than accuracy of the two systemswork independently.This article focuses on the increasing the number of sensors of IntegratedPositioning System to obtain more accurate positioning information. but the hardwarerequirements, high cost, engineering implementation difficult,so this paper uses acombination of loosely coupled witch low cost, suitable for civilian. Loosely coupledsystems of higher DR sensor requirements, and the original DR system includes onlythe odometer and rate gyro two ordinary sensors, in order to achieve a more preciselocation information, This article will add affordable, live usual accelerometer andgeomagnetic sensor in turn to the Integrated Positioning System, the formation of threeCombined models with different sensors, namely “GPS/DR integrated model”,“GPS/DR/Accelerometer integrated model” and “GPS/DR/Accelerometer/Geomagneticsensor integrated model”, so that each sensor can give full play to their advantages,provide a more complete location information for the system, thus improving the overallpositioning accuracy.This paper studies the Kalman filter method in combination positioning,and usesthe unscented Kalman filter algorithm witch can effectively deal with non-linear systems to fuse the data from different sensors. In this paper, using Matlabprogramming software for the three models were simulated and analyzed, the resultsshowed that the combination in a loosely coupled manner, increase the number of usefulsensor helps to improve positioning accuracy of Integrated Positioning System.