Optimal Game Theory in Complicated Virtual-modeling and CGF Decision-making with Multi-granularities

Computer generated forces (CGF) with multi- granularities and complicated nonlinear virtual natural phenomenon modeling are all key technologies in the field of Distributed Interactive Simulation. An ´optimal game theory´ was proposed for the first time to establish special modeling and nonlinear decision-making, some principles and algorithms were also deduced from this theory. The original inspiration of this theory was to optimize crucial factors such as complexity of physical modeling, computing efficiency, bottleneck of human sensory organ, granularity of simulation and make an effective virtual simulation scheme relatively. Junior and senior models derived from the main principles were respectively used to simulate models with multi-granularities and complicated decision-making problems. Crucial equation appending random disturbance holds the uniform scenario of Junior. A virtual 3D system of dynamic stars, moon and their mirror images in water was presented according to the junior model. Efficient texture blending and particle system made the crucial equation workable while disturbance algorithm made the stars twinkle more vividly. Variables with different weight participate in main equations. The output gives an optimal result in senior model. An armor CGF with multi-granularities is provided to testify the feasibility of senior model. Damage-summarizing and dynamic algorithm of aggregation was discussed using generalized mathematics expectation and fuzzy mathematics. Finally, an intelligent observer in large scale 3D battlefield was designed using synthetic model. Quantitative parameters of nature, sense and complexity were presented and weighted observational mode equation was calculated by generalized expectation which made the virtual scenes more smoothly and floridly.