The Crowd Simulation Under The Traffic Environment

The crowd simulation is the use of artificial intelligence, computer graphics technology, through the establishment of the feature model of crowd behavior, to complete the simulation of crowd movement in the real world. The crowd simulation technology is an important part of virtual reality and computer simulation technology. The applications of crowd simulation include:various types of virtual environments, building evacuation assess, group behavior walkthrough, video production, computer games. The crowd simulation algorithm has the following features:a high degree of complexity of the algorithm, the computational overhead and real-time requirements. When the crowd reached a certain number, only rely on the CPU serial mode line crowd simulation, it is difficult to meet real-time requirements. In recent years, with the rapid development of GPU technology, GPU has a strong computing power, especially in the parallel computing capabilities very prominent, very suitable for crowd simulation.Based on the study of character animation algorithm, crowd simulation algorithm, GPU computing technology, this thesis, introduces GPU-based computing technology to the field of crowd simulation. The main contents of the thesis work include:Design a skeletal animation rendering GPU-based parallel algorithm. In the crowd si-mulation process, the traditional role model rendering will consume a lot of CPU resources. This thesis analyzes the characteristics of the role of skeletal animation, design a parallel computing algorithm of the skeletal animation, using the OpenGL Shading Language (GLSL) to achieve its optimization. The algorithm allows GPU to complete the bone matrix opera-tions. The specific method is:pass the bone matrix in the form of texture data to the GPU vertex shaders and use the vertex shader to complete the deformation operators of skeletal animation mesh. The algorithm takes full advantage of the GPU parallel computing technol-ogy and it’s a better solution on the calculation of crowd simulation which based on charac-ter animations.This article proposes a crowd behavior policy model based on the Signal Vector. The person’s behavior strategy is established on a signal system, using the signal vectors to drive the behavior of the individual person. The model uses the signal vector to describe the ob-stacle collision information, environmental information between people. The model simpli-fies the computing of the behavior of the crowd which is able to meet thousands of individu-al characters to move freely in the scene, and to avoid the collision between people and people with obstacles. The model facilitates the expansion and optimization of parallel computing.Based on OSG (Open Scene Graph) three-dimensional graphics engine, we build a set of virtual traffic environment platform to subway platform environment. Based on the virtual traffic environment platform, we realized the crowd simulation algorithm described above, and carried out the related experiments. The experimental results show that the method used in this paper is feasible and better to improve the efficiency of the crowd simulation, which is to ensure the real-time simulation.