Research On Query Spatial Objects With Uncertainty

With the development of spatial information technique and further research of the theory of spatial data with uncertainty, uncertain data have been concerned widely. As the data with uncertainty are widespread, people paid more and more attention in the database field.The ordinary database management technology has been unable effectively to manage the data with uncertainty. Therefore, many domestic and overseas researchers exploit management technologies of the data with uncertainty as one of the important research of the spatial database with uncertain data. Query, analysis and processing are one of the key technologies of the uncertain database management, but the study of special objects with uncertainty query are less. In the paper, we discuss that spatial objects abstracted to points or lines with uncertainty under several different environment.As the input data with uncertainty, the results are often approximate, how to make query results more precise and reasonable is the main starting point of our study. In this paper, the storage, index and query of the data with uncertainty are further studied. The uncertainty of the spatial objects appear in the inherent uncertainty of spatial objects self and relationships among spatial objects.At first, the inherent uncertainty of spatial objects is studied. As the index of the future trajectory and query with uncertainty, the technology of the trajectory index and query of spatial moving object can be as a starting point. We researched the full-time indexing and query of the past, present and future information of the moving objects on the road network. Finally, the segments with uncertainty are further studied, including the spatial relationships of segments with uncertainty. Finding included the following aspects:L2R tree index is proposed, which consists of two layers of R tree and a linklist. The two-level R tree index for road networks and the past movement information of moving objects, the objects’current position and the information of future predicted trajectory are stored in the linklist. This index can effectively carry out a full-time query for moving object and based on this index achieved moving objects range queries and point queries. A full-time indexing structure based on a simulation prediction model in city networks is proposed. According to the characteristics of urban traffic network, using a cellular automata model with crossings to simulate future trajectories of moving objects, and using linear regression and circular arc fragmented curve-fitting method to obtain the trajectory prediction equation of moving objects in the regular route and in crossing respectively. Experiments show that the query and update algorithm of AUC index is efficient.Based on the uncertain trajectories model existed, a method to model the trajectory of moving objects in road network is proposed.Based on this model, a probability means used to achieve the point queries and range queries for moving objects with uncertainty, and the experiment shows that the probabilistic method can be carried out with better performance.A new Nearest Neighbor query method of the moving objects with uncertainty in road networks trajectories is proposed. An existed uncertain trajectory model is adapted when nearest neighbor query is performed with uncertainty considered, and the nearest neighbor query method with probability will be applied to nearest neighbor query of moving objects with uncertainty in road networks. A method of networks influence region gradually expansion of query object is proposed to compute the probability that the candidate object will be the nearest neighbor of given query object, and the probability value as the nearest neighbor query results. The proposed method can realize the nearest neighbor query of moving objects with uncertainty in road networks.The subject of segments with uncertainty nearest neighbor query is proposed. The method expressed the fuzzy region of the segments into the border area with uncertainty, and the affected area of the segments is considered.The distance between two line segments is calculated according to their position relationships separately. The probability formula is used to calculate the impacted degree and the purpose is to measure the impacted level by specified line segments.