Study And Implementation Of Vehicle Navigation System Based On Tightly-coupled GPS/DR

As an important part of intelligent transportation system (ITS), intelligent vehiclenavigation system is playing a significant role in improving the operational performance ofthe modern transportation system. However, the traditional navigation based on GPS hastremendous difficulty in satisfying the stricter requirements for navigation performance.Based on new requirements of this kind of social contradiction for vehicle navigation, GPSand DR tightly-coupled navigation algorithm is designed in this paper. Meanwhile,Map-Matching (MM) with higher success rate has also become one of the important issuesthat need to be addressed by the intelligent vehicle navigation system in the urbanenvironment. A map-matching algorithm is designed with multiple reference information suchas speed, heading of the vehicle, the connecty of the links and so on. In order to run themap-matching algorithm, a map data struct is designed in this paper with space index struct.On the other hand, so as to ensure the quality of the vehicle navigation system, an enhancedReceiver Autonomous Integrity Monitoring (eRAIM) algorithm based on GPS/DR/MMtightly-coupled is designed in this paper. This paper takes research in the following aspects:1. Analyze GPS/DR tightly-coupled integrated navigation algorithm.Firstly, the model of single point positioning based on GPS pseudorange and the errorssources which affect the GPS positioning accuracy are analyzed. Then DR inertial unit wasintroduced, analyzed the the observations and theory of navigation. Because of the GPS/DRloosely-coupled navigation failing to meet the positioning accuracy requirement and theobservation equation being nonlinear, the tightly-coupled integrated navigation system whichfused the GPS and DR measurements based on Extended Kalman Filter (EKF) are designed.2. Research on map data storage structure.In order to reduce the physical storage space occupied by map data, a data structurebased on space topology for electronic map data is designed in this paper. With this structure,data and attribute information of electronic map can be stored with the proposed space indexstructure. In this paper, the map data is stored in integer data and without loss accuracy of data.In order to achieve the coordinate translate from ECEF to local plane coordinate, a newmethod is designed in this paper, because this method achieve coordinate translate efficientlywith six parameter, the method called “Six Parameter” method.3. Research on Map Matching (MM) algorithm.With the map data stored in the storage and the positioning results from the tightly-coupled algorithm, a new method is put forward to search candidate links, whichachieves with a rectangle search range, namely rectangle window method. Considering thespeed of vehicle, the heading of the vehicle, the distance between the positioning point to thematched point and the connection of the map links in complex, a map matching algorithmbased on point-to-line matching method designed.4. Research on RAIM algorithm based on GPS/DR and map matching.With the innovation which comes from the process of the enhanced Kalman filter forGPS/DR/MM and the variables of the probability of miss leading and fault alarm, anenhanced receiver autonomous integrity monitoring algorithm is carried out. With theenhanced RAIM monitoring the errors from the measurements of GPS and the faults of thesatellite signal and the fault map matching.5. Design the hardware platform of vehicle navigation system with ARM.With the thought of object-oriented, the software for data processing based on Linuxsystem is designed. With two serial ports collection measurements from GPS and DRrespectively, the hardware platform for vehicle navigation system with ARM as the control ofthe core is build. All the algorithms such as tightly-coupled navigation algorithm, mapmatching algorithm and receiver autonomous integrity monitoring algorithm will run on theplatform. At last, the terminal display function is designed for the navigation system with atouched LCD, witch can achieve Human-Computer Interaction to set different modes.In order to validate the theoretical analysis results, filed data collection was carried out inoutdoor environment, experiment results demonstrate the feasibility of the proposedalgorithms. Compared with the standalone GPS, the result shows that the loosely-coupled andtightly-coupled navigation algorithm increased the positioning accuracy by50.8%and68.3%respectively. The experiment shows that the GPS/DR tightly-coupled integrated can be anavigation algorithm for urban vehicle. By contrasting the performances of differentnavigation algorithms, it shows that the RAIM algorithm can monitoring the fault from GPSsignal and mapmatching. Through the map matching real-time correction error in thenavigation process, the experimental results show that the success rate of the map matchingwith the positioning result of the tightly-coupled algorithm can be above99%, meeting therequried navigation performance that civil aviation non-precision approaches require.