Implementation of GPS / INS Fusion Algorithm Using GPS Pseudo-Range

The navigation system of vehicles calculates the speed, position and attitude of the moving device relative to a reference frame and provides it to the guidance system. One of the most widely used navigation systems is the inertial navigation system. Due to the increasing error of the navigation system over time, the integrated navigation system is usually used for long-term navigation. One of the most common integrated navigation systems is the INS integrated navigation system with GPS, each of them which has advantages and disadvantages that cover the other. In this paper, two GPS and INS data integration algorithms with loosely and tightly coupled integration are implemented and compared. In the loosely coupled method, GPS measurements include positions and speeds. In the tightly coupled method, a model for GPS error is considered, which includes bias dynamics and GPS clock drift. The result of combining GPS and INS data in this way is closer to the truth, but in the method of loosely coupled, the result of the combination follows the average of GPS data. In the implementation of the combined algorithm with tightly coupled, raw GPS data is used, which is pseudo-range and pseudo-range rate along with astronomical information. In this paper, an extended Kalman filter is used to integrate the data of two measurement data. The simulation results show the superiority of tightly over loosely connection performance. Also, the integration algorithm with a loosely approach has been implemented in hardware and car testing has been done in two scenarios of connecting and disconnecting GPS.