Research And Implementation Of Query Technologies Of Moving Objects In Transportation Networks

Transportation network databases, which are on the basis of spatial databases and spatio-temporal databases, mainly study the objects moving in constraint environment. The core is the problems of the modeling of networks and modeling, index, query of the moving objects in networks. The aim of researching moving objects in networks is to provide a reliable platform of artificial transport system for transportation analyzing and decision-making. This paper focuses on query processing technology of moving objects in transportation networks, the main work are as follows:Firstly, the paper introduces the current development of transportation network database, discusses the concept and the characteristic of mobile objects of transportation networks, describes in detail modeling of the transportation networks, the types of modeling and index of moving objects in networks, and analyzes the application background and methods of query.Secondly, for the issue of k nearest neighbor query in transportation networks, using storage structure of supporting transportation networks connectivity information, the paper proposes PM-KNN search algorithm to solve k nearest neighbor query in transportation networks .It bases on m neighbors of the important point which is pre-computed. This method and island method are different. This method does not calculate the point around the points of interest, but the operator can choose to use part of important point in networks. As long as the transportation networks structure constant, the cost of pre-computed points of interest will not increase.Thirdly, for the issue of reverse k nearest neighbor query in transportation networks, algorithm use pre-computed KNN query results, proposes pruning method of the query space, proposes the reverse k nearest neighbor query algorithm PM-RKNN based on networks expansion, reduces the access of nodes and the points of interest.Fourthly, for the issue of group nearest neighbor query in transportation networks, distribution range of the target object and the query object may be dense or sparse. This paper proposes to use two different pruning strategies according to the distribution range of the target objects and the query objects, proposes an improved group nearest neighbor algorithm CMBM.