Distributed Method Of Road Planning And Design Of Prototype-system Based On Hybrid Vector Data Structure

The definition of road planning is that according to the demands of engineering and geographical information such as geology, topography and terrain, planning route for road design and construction.The existing method of road planning is mainly based on the geographic information system, by using traditional centralized storage, stand-alone computing and grid data structure etc. technology. However, with the development of digital engineering design, especially the explosive growth of data size, current method of road planning is difficult to satisfy the needs of engineering data processing.Recently, with the improvement of cloud computing technology, engineering road planning based on the big data is possible. Therefore, in this paper, according to the idea of Big Data, research the method of road planning based on hybrid vector. This technique uses a hybrid vector structure to describe geographic data, proposes a pre-processing method based on the distance of clustering of geographic data. Hybrid vector structure is an extended geographic data structure of current vector structure in terms of type and description of data, which designed according to features of geographic data and NoSQL database. The pre-processing method based on hybrid vector structure converses attribute of geographic data at first, and then according to the idea of clustering, compresses and classifies geographic data. At last, calculate the geometric center, area, boundary property of subset region. Using the results of pre-processing, and according to the distributed method, this paper proposes the method of road planning based on the constraint of control-line. According to the control-point which is the area that road must cross through, and the connectivity between control-point, control-line can be determined. By dividing control-line to control-line-fragment, and utilizing control-line as the constraints of distance and direction, the problem of road planning in a multiple control-point area transform to the road planning between two points. Referring to the attributes of subset area which contain a control-line-fragment, proposing a comprehensive evaluation method to solve road planning problem between two points, At last, by combining the solution of control-line-fragment, get the result of road planning in multiple control-point area.In order to validate the proposed method, this paper adopts Hadoop cluster technology as the distributed computing, storage architecture, using Java and JSP and MyEclipse development environment, design and implement the prototype system of road planning. In the end, using geographic information data of a valley area to verify the proposed method of road planning, prove the effectivity of the proposed method.