Research On The Key Visual Technology In Large Scale Warfare Simulations Platform

The large-scale combat simulation is an important means for military training and combat in the world. Visualization technology is the key method that simulation approaches realistic combat. Along with the training scale becoming larger and wider, the entities participated being more, the 3D visualization comes to be a prominent trouble in large-scale combat simulation.The visualization technology in large-scale combat simulation such as retrieval of the 3D model, LOD rendering of the battle field, the fast rendering of colony, collision detection of models are deeply researched. The relevant technology solutions are put forward. They provide creative approaches and methods for visualization in large-scale combat simulation.1. A 3D model matching method based on flattened surface is presented. The flatness-based surface flattening principle is analyzed, which can be applied to flatten the surface of model into a 2D shape as well as automatically cut and partition. The method avoids global and local self-intersections and minimizes the total length of the introduced seams. Then a shape matching methodology based on Fourier Transform is studied. By using the flattened surface, pole map version normalization and Fourier Transform, the shape similarity comparison is performed. Another TAR-based shape matching methodology is designed and implemented. With areas of the triangles, the convexity of each point of flattened surface is measured at different scales and the shape similarity comparison is performed.2. A fast nine palaces terrain rendering algorithm is put forward. The multiresolution ground block structure is set up with characteristic of nine palaces. Level switching of resolution occurs within 8 surrounding regions. The operations of dividing terrain and viewpoint center squared are combined to one and the dividing number is decreased. At the same time, to avoid cranny bought by multiresolution model, the cranny removing algorithm is put forward. A novel visibility algorithm based on progressive horizon is proposed. With the contour of models, the progressive horizon can be constructed and occlusion culling is accelerated. The performance analysis and error estimate of generated terrain are studied. A mobile agent-based large-scale distributed battlefield environment scheduling framework is presented. It has the capability of scaling across geographically distributed resources. The architecture consists of distributed mobile Agents working cooperatively to support and manage the distributed collaborative virtual environments.3. A fast based-SPH (Smoothed Particle Hydrodynamics) colony rendering method is proposed. The colony is represented both discrete individual and a single continuous system. The task-based behaviors planning scheme and individual autonomous behaviors model are studied. A novel bidirectional distance constraint method based on continuous dynamic is exploited to render the large-scale behavior of the colony. It accelerates individual collision and avoidance in the movement of the colony.4. A collision detection method based on topology level graph is developed. The method of constructing topology level graph is disposed. The model is divided into different sub-objects. Then the sub-objects are looked as roots to build OBB level trees. A* method is improved to fit searching topology level graphics, and the triangle-triangle intersection test method is modified to reuse intermediate computation results. The operation number is decreased in the process of testing two OBB leaf nodes.