Research On Model Of Transportation Network Databases And Key Dealing Technology Of Moving Objects

With the development of database technology, transportation network databases as one of the most important branches of moving object databases, has become increasingly wide range of applications. At present, the study of transportation network databases has many problems in date model, storage structure, query, and index and so on. This thesis mainly studies the road network modeling, the operation algorithm, and the location update and predication strategies of moving objects and so on. The main research work is divided into the following three parts: Firstly, the thesis analyzes the shortcomings of route-oriented model which proposed by professor R.H.Güting. It proposes a data model named NHTND which based on cross and roadway.First, the model uses roadway as basic of specific location which expresses the road network more detailed. Afterwards, it optimizes the express of road and cross, which performances better than route-oriented model. The model supports for the traffic lights and three-dimensional crosses, meets the needs of dynamic properties in urban road network. Finally, the model expresses the moving objects with velocity and acceleration, which raises the precision of expression and guarantees history, now, future query.Secondly, according to the phenomenon and need of real traffic, with introduction of cellular automata theory, the thesis proposes the operation algorithms of NHTND model. It analyzes the micro-relations of moving objects and proposes mobile relational operations. The set has seven basic operations that will do much good to the movement of moving objects, prediction and analysis. Based on the classification of the jam, NHTND proposes the algorithms for forecasting regional drift, dead time, which is the basic of optimal path with barriers algorithm. All of the algorithms improves the operation algorithm of the model.Thirdly, aim at the problem of moving objects'location update and predication strategies in the research, the thesis analyzes mobile law of moving objects in real traffic and proposes a location update strategy named AVD_LUS, a location predication strategy named AVD_LPS, which are based on the acceleration. At the same time, it expands the meaning of Where_At query and proves the results. The experiment shows that the strategies effectively reduce the location update times as basic, improve the query accuracy and guarantee the effective tracking of moving objects.