Research On Technologies Of Intelligent Vehicle Integrated Location And Route Navigation

Vehicle location and navigation system, which belongs to the research category of the Intelligent Transportation System (ITS), is an important ITS application system in great demand, and is also one of the efficient method to solve the urban traffic problems recognized by the international. Vehicle navigation system uses synthetically the detection, communication, computer, control, GPS, GIS and so on to supply the vehicle position, the real traffic information, route plan and route guidance to the drivers dynamically. It made traffic control centre site the position of vehicle truly, prevent the traffic block and reduce the air pollution.In order to effectively carry out the development and research work of vehicle location and navigation system, this thesis is all about the theory model and key technique of vehicle navigation system, in which the principle is that the science significance is equal to the practical value, and make a series of innovation fruits on the researches of information fusion of vehicle location, map matching, road impedance evaluation, the dynamic route optimal technique, the three-dimension navigation techniques, the design and operation technique of information inform in vehicle navigation system and so on. The major work in this thesis can be summarized as the following:GPS/DR integrated location system is studied in order to improve the location precision, and ensure location reliability and continuity. The measurement model and state model of GPS/DR integrated system are constructed, and the integrated design scheme based on federated Kalman filter is studied. An adaptive Kalman filter algorithm based on fuzzy theory is used in the system. This method changes measure noise covariance matrix R on-line based on the actual measure data and it can improve the filter precision. A fuzzy weight reasoning system is established for many minor filters to fuse their information. This method lessens the main filter calculation and improves the positioning accuracy of the integrated navigation system. Because of navigation system complexity and the uncertain surroundings, there can't build veracious fuzzy system through human experience. The genetic algorithm is used to optimize fuzzy membership function and obtain the optimal or sub-optimal control rules. The results of simulation demonstrate the feasibility and effectiveness of the method.The map-matching method, through which the digital map information can be used to calibrate the vehicle location estimation provided by positioning system, is systematically analyzed. Various issues that associated with the correctness of map-matching calibration are discussed, and an original map-matching algorithm based on the probability decision rule and the D-S evidence reasoning theory is proposed. According to the vehicle's current running status, the basic probability assignment functions about vehicle's position and direction information are designed. According to the D-S combination formula, the fusion of position and direction information is made. The pignistic probability transforms for decision making is introduced to get better performance evaluation. It assigns smaller values of probabilities for smaller values of Beliefs while assign higher probability values for higher values of Beliefs which lead to better decision making. Simulation results with actual field experimental data show that algorithm can perfectly solve the MM problem at urban road net.Shortest path algorithm is the basis of a correct and effective vehicle navigation function. In this thesis, the systemic method of shortest path planning in a vehicle navigation system is studied. An evaluation method of route synthetical impedance based on traveler act characteristic is proposed. Based on the continuous property of the traffic flow, a dynamic discrete-time road network is established, an effective dynamic adaptive routing algorithm is proposed based on the idea of A* algorithm. In order to overcome the disadvantage of conventional shortest path algorithm, we proposed a method based on virus evolutionary genetic algorithm to solve the optimal path problem on vehicle navigation system. By applying random A* algorithm, the initial generation of genetic algorithm is obtained. The local search by horizontal propagation of evolutionary information in the same population with virus infection operators is made, so it can get the satisfied solution in less time than genetic algorithm. This algorithm can quickly find the optimal path and does not need any network constraint condition, which also can solve the problems on continuously and discrete networks. The road traffic network of urban is more and more complex day by day, the existing two-dimensional electronic map isn't able to reflect the staggered relations of the roads direct-viewing and the vehicles navigation system based on the two-dimensional electronic map appears helplessly facing the route guiding of the complex crossings. The dissertation studies the route navigation technique in vehicle navigation system, establishes three dimensional virtual scenes of the complex crossings based on Multigen Creator and realizes the scene roaming of the assigned route based on Vega. The dynamic response between the two-dimensional electronic map and the three dimensional virtual scene is realized in case of overpass. It will fully combine the entirety, macrograph of two-dimensional electronic map and the direct-viewing understandability of the local three dimensional virtual scenes for enhancing the guidance efficiency of the existing vehicles navigation system.Lastly, the design project, hardware and soft composing, function introduce of vehicle navigation are studied in the thesis.