Research On Real-time Dynamic 3D Visualization Of Abstract Data Relations

Visualization of abstract data Relations is aimed mainly at visualizing data structure, and graph is the most general and frequent used data structure. Visualization of graph including static and dynamic graph visualization. Since dynamic graph can be regarded as a series of static graph, static graph layout algorithm serves as a basis for dynamic graph.Layout drawing for general static undirected graph is transformed into an optimization of certain objective function modeled to reflect certain aesthetics of graph layout. A solution for this numerical problem brings optimized graph layout. And layout drawing for dynamic graph is usually updated by the same static graph layout algorithm. So the result of layout drawing is lack of contact between graphs, and has made it impossible to maintain the stability of dynamic graph.This paper applies Coevolution Genetic Algorithm in graph layout algorithm, where coevolution serves to solve the optimization of objective function. As for dynamic layout, applying "Jump the moving queue" in layout algorithm for dynamic graph based on Coevolution Genetic Algorithm is aimed to maintain its stability.Coevolution Genetic Algorithm for Graph Layout divides the whole population into two subpopulations which communicate with each other in the process of evolution, simulation process for the purpose of achieving better or faster evolutional effect than in counterpart case of only a single population. Experimental data has shown that Coevolution Genetic Algorithm applied in graph layout of general undirected graph can effectively overcome the shortcomings of simple genetic algorithm, such as a tendency of local optimum, or a significant reduction of convergence with the increasing scale of the population etc. Further, in respects of convergence and temporal cost for it, Coevolution Genetic Algorithm for Graph Layout is superior to Genetic Simulated Annealing Algorithm for Graph Layout.The initial layout of "Jump the Moving Queue" stability layout for dynamic graph is produced by Coevolution Genetic Algorithm for Graph Layout, so the aesthetics of dynamic graph are determined by the aesthetics of Coevolution Genetic Algorithm for Graph Layout. In this algorithm, graph layout is smoothly updated according to progressive topological changes. By controlling barycenter and pinning weight defined in this paper, the update effect of "large partial changes, subtle overall changes" is achieved. And, positions of vertices are managed both by structure and changes of the updated graph, the algorithm essentially take the whole undated structure and continual stability into account.These two algorithms are tested respectively and are proved respectively their feasibility and effectiveness in graph drawing for static and dynamic graph.