Research On Real-time Cloud Dynamic Simulation Algorithm Based On Cellular Automata

The simulation of natural scenery is a research difficulty in virtual reality and system simulation.As an important fuzzy scene,cloud can enhance the realism of virtual scenes,and has a very wide range of applications in weather prediction,film and television special effects,flight simulation and computer games.Clouds have irregular shapes,unpredictable changes,and complex optical phenomena,making it difficult to describe using traditional Euclidean geometry methods.Therefore,how to simulate realistic 3D dynamic cloud in real time becomes a challenging topic.In recent years,with the rapid development of computer hardware,the simulation technology of the cloud has continued to deepen.Researchers at home and abroad have proposed many simulation algorithms through continuous research and practice.They are mainly divided into physical-based methods and heuristic-based methods according to different modeling techniques.The physics-based approach mainly considers various factors related to cloud formation and simulates the natural formation process and change process of clouds.This kind of method can generate realistic clouds,but it is difficult to achieve real-time performance when applied to a large-scale virtual scene because its principle is complex and the amount of calculation is large.The heuristic-based method mainly uses the visual morphological features of the cloud to generate the density distribution of the cloud approximately,and simulates the cloud by generating a texture.This method has a small amount of calculation and is easy to implement,but the generated cloud lacks details and is not very realistic.This paper first describes the characteristics of three-dimensional clouds in nature and the complexity of simulations determined by their characteristics,and classifies them according to certain criteria.Several commonly used cloud simulation methods and current research status has been discussed,and their advantages and disadvantages has been analyzed respectively.The physical formation process of the cloud and the fluid dynamics principle are analyzed,and the solution method of the fluid equation is given.Combining with the characteristics of cloud,the theory of cellular automata is deeply studied.Three-dimensional dynamic cloud is modeled based on this,and its generation,dissipation and wind movement effects are simulated.By setting the ellipsoid and its parameters in the simulation space to control the shape of the cloud,different types of clouds are generated.For the complex calculation problem in the smoothing process of the cloud,the Sigmoid functions are used to generate a continuous density distribution of the cloud.An improved multiple forward scattering illumination model is proposed to reduce lighting calculations during rendering.GPU-based ray casting algorithm in volume rendering technology have adopted to accelerate the rendering of clouds and generate a realistic 3D cloud.Finally,the simulation experiment of this algorithm is carried out.By controlling different parameters,the simulation of cirrus cloud,cumulus cloud and stratus cloud is realized,and the illumination effect of cloud can be freely adjusted.The experimental results show that the algorithm is better in real-time and more realistic,comparing to texture mapping,particle system and fractal geometry methods.