Parallel space topology analyzing method based on discrete grid

Abstract

The invention discloses a parallel space topology analyzing method based on discrete grid. In the method, data is loaded firstly during the parallel space topology analyzing method. Then, the data division is carried out on a space object by Root process. Divided sub-data are transmitted to a plurality of Sub processes. Every Sub process carries out the space topology analysis on accepted tasks in parallel. The parallel space topology analyzing method has perfect topology robustness. Data communication is reduced. Compared with a serial method, the efficiency of the parallel space topology analyzing method is improved linearly.

Description

technical field [0001] The invention relates to a parallel space topology analysis method, in particular to a discrete grid-based parallel space topology analysis method. Background technique [0002] The characteristic of spatial data is that it not only includes the spatial position and state information of the object itself, but also includes the information representing the spatial relationship of the object. Spatial relationships mainly include metric relationships, directional relationships, and topological relationships, among which topological relationships are the most important and complex, and also contain the most information. Topological properties are the properties of spatial objects that remain unchanged under continuous deformation, such as containment relationship, adjacency relationship, etc. In topology, there is no element that cannot be bent, and the size and shape of each figure can be changed. The study of topological relations originated from the p...

AI Technical Summary

Problems solved by technology

[0005] (1) Spatial topology relationship analysis is frequently used in database operations such as spatial query and update, and topology operations are CPU-intensive operations, which are very time-consuming. In spatial databases with massive data, the response to topology analysis operations Time does not fully meet the user's requirements for real-time response

[0006] (2) The current hardware development trend is to develop cheap parallel computer systems based on traditional memory, processors, and disks to handle complex calculations. It is clear that such hardware requires parallelization of spatial database operations to be able to play. utility

[0009] (1) Parallel Processing of Spatial Joins Using R-trees by Thomas Brinkhoff, Hans-Peter Kriegel and Bernhard Seeger in Proceeding of the 12th International Conference in 1996 and by X Zhou, David J Abel and David Truffet in 1998 Data Partitioning for Parallel Spatial Join Processing on GeoInformatica does not divide spatial objects, but divides different spatial objects to different disks, and does not divide a single spatial object. If a single large spatial object is used according to the method After the division, the topological relationship between spatial objects will be wrong or missing;

[0010] (2) The types of existing spatial topology analysis operations are incomplete. For example, Efficient Parallel Spatial Join Processing Method in a Shared-Nothing Database Cluster System published by Warnill Chung, Soon-Young Park and Hae-Young Bae on LNCS 3605 in 2005 has no space The results of data division are saved, and the division results cannot be reused. Instead, the spatial objects are only divided during operations such as Join for parallel processing.

Images