Research On Representation Of Radar Coverage In Virtual Battlefield Environment

In modern war, the description of electromagnetic environment is a key factor which restricts the electronic counterwork commander to command well, truly and efficiently. Whether to get predominance in electromagnetic control not only depends on performance of the equipments, but also on nice judgment of electromagnetic situation and reasonably deployment of electronic counterwork power. Radar is one of the most primary elements which affect the electronic environment in battlefield. How to represent the Radar effect vividly to commander and drillmaster, especially the effect under complex environment, so can they hold advantages and discover problems, then generate more profound comprehension and apperception, accordingly hold entire warfare from higher level and hold the battlefield situation in time, becomes a crucial problem in the future Digital Virtual Battlefield Environment.Traditional manners of representing Radar coverage are mostly in 2D, which are not intuitive or agile. For different application requirements, this thesis researched two methods to construct 3D Radar coverage model separately based on Radar equation and electromagnetic wave propagation. Furthermore algorithms related to model constructing, model simplifying and rendering were designed and implemented for higher efficiency and fidelity.In the process of theory researching and algorithm practicing, based on contents to be researched and problems to be resolved, some creative algorithms and methods have been proposed. In detail, the highlighted ideas and main contributions of this thesis are described as follows:For reference, five steps of data filed visualization in 3D space such as datum generation, datum refinement and process, visualization mapping, rendering and displaying were used. A framework of 3D Radar coverage representation in virtual battlefield environment was put forward according to following four steps: data field building, data preprocess, 3D model generation and integrated representation.The advantage of digital terrain data provided by virtual battlefield environment was taken to estimate the maximum range Radar can detect based on Radar equation. Based on the 3D representation of Radar coverage in free space, according to geometric optic method, the influence of a single peak was investigated, and then an algorithm was proposed to represent the 3D Radar coverage affected by single peak. Then 3D coverage influenced by continuous peaks was get. This algorithm is best applicable for cosmically virtual battlefield environment which requires relatively low precision.Started from electromagnetic wave propagation, propagation loss on grid nodes in plane was calculated based on APM. Then a virtual 3D strategy was designed to combine multi in-plane loss value to construct a 3D data field. From the data field visualization point of view, an algorithm of construct maximum detection coverage model was proposed based on 2D contours. For representation of details in the coverage, another model construction algorithm was put forward based on isosurface extraction.For relieving the burden of CPU, started from the simple tetrahedron case, hardware accelerated isosurface extraction algorithm of hexahedron case was deduced. Then the process of isosurface extraction was moved from CPU to GPU. After a series of key problems such as datum preprocess, in/output mapping to texture and isosurface extraction algorithm, etc were resolved, hardware accelerated 3D Radar coverage model construction was implemented.A view-dependent mesh dynamic multiresolution data structure was designed which can support topological structure modification. In order to produce high fidelity and drastic simplifications of complex models, the algorithm needed to combine disjoint parts of the models. So the general vertex pair contraction operation was used which could collapse vertex pairs that were not connected by an edge. The number of candidate vertex pairs was limited by adaptively selecting the distance threshold and defining a unified error metric that combines both geometric and attribute error to control the order of simplification. Based on the sequence of vertex pair contraction operations, by reorganizing the simplified data a view dependent multiresolution data structure was built, so could the proper detailed simplified model be generated or the rapid transformation between different LOD models be implemented according to the view point parameters.A prototype system of 3D Radar coverage representation in virtual battlefield environment named 3DRCRepresent was designed and implemented, which gave a sound support to our 3D Radar coverage representation framework in virtual battlefield environment and related techniques. And some parts of this system were applied in related projects.