Design And Research Of Large-Scale Scene Visual Simulation Based On LOD

Simulation technology is the technology that has a science experiments on system model using computer. It has the advantages of economic, practical, safe, flexible and repeatedly used. Simulation technology has applied in all fields of national economy currently. At the same time, computer simulation technology is urging science research to develop effectively as the kernel of virtual reality with the first guidance of information technology and computer technology. The basic purpose of simulation reduced to decision-making event. Making GIS 3D visualization are very important practical significance in the military, simulation, virtual reality, earth science, and other disciplines.After researching all kinds of definitive approach of virtual reality system, this paper sets out to resolve the bottleneck in very large terrain data visualization, propose one method based on virtual reality LOD (level of detail) that include HLA (high level architecture) and parallel computing, and design a simulation system of large-scale water conservancy projects scene. Have been realized some of the visual scene using visual C++ and OpenGL.The main work and result include following parts:1. Introduce the application and technical background of LOD, summarize and review the characteristic, research condition of irregular networks LOD.2. Discuss the technology question about large-scale terrain data parallel compute based on irregular networks, terrain triangulated irregular, model generating algorithm and reasonable calculational methods.3. Compare several of the representative irregular networks LOD algorithms.4. Put forward design philosophies that integrate virtual reality, parallel computing technology and high level architecture for enabling visualization more efficient on PC.5. Using visual C++ and OpenGL to exploit a simulation system, and have an experiment about large-scale terrain real data. The results show the design method in this paper is feasible.