基于物理的布料模拟研究综述

Review of Physics-Based Cloth Simulation

中文关键词: 计算机图形学  物理仿真  布料模拟  布料动画

英文关键词: computer graphics  physical simulation  cloth simulation  cloth animation

中文摘要:

      目的 在图形硬件计算性能不断提高，物理仿真动画技术蓬勃发展的背景下，对基于物理的布料模拟技术中的主流方法进行总结和归纳，指出未来发展趋势，为相关技术与应用研究提供参考。方法 通过文献收集和整理，从布料物理模型、物理模拟算法、碰撞处理算法和高分辨率加速算法等方面入手，对基于物理的布料模拟相关研究成果进行综述，分析现有方法的本质、优缺点和内在联系，基于分析归纳预测其未来发展趋势和所面临的挑战。结论 总结归纳了布料物理模拟算法的研究进展，研究围绕隐式时间积分无约束优化问题展开的共同点，以及GPU并行加速和多重网格方法对于高分辨率布料实时模拟的贡献，指明了布料物理模拟的发展趋势是模拟动画真实感与模拟算法高效性会得到不断提高，其面临的挑战是随着硬件性能的发展进一步提高模拟精度，以及在有限计算性能下需要更好地平衡模拟精度和效率的问题。

英文摘要:

      In the context of the improvement of graphics hardware computing performance and the development of physical simulation animation technology, this paper aims to analyze and summarize the mainstream methods of physics-based cloth simulation technology, point out the future development trend, and provide a reference for related technology and application research. Through literature collection and collation, this paper summarizes the research results of cloth simulation based on physics from the aspects of cloth physical model, physical simulation algorithm, collision processing algorithm and high-resolution acceleration algorithm, analyzes the essence, advantages and disadvantages and internal relations of existing methods, and predicts its future development trend and challenges based on analysis and induction. The research progress of cloth physical simulation algorithm is summarized, the common points of the research around the implicit time integration unconstrained optimization problem are studied, as well as the contribution of GPU parallel acceleration and multigrid method to high-resolution cloth real-time simulation. It is pointed out that the development trend of cloth physical simulation is that the realism of simulation animation and the efficiency of simulation algorithm will be continuously improved, the challenge is to further improve the simulation accuracy with the development of hardware performance, and to better balance the simulation accuracy and efficiency under limited computing performance conditions.