基于自适应Runge-Kutta法的织物动态模拟

Cloth animation based on adapted Runge-Kutta method

运动微分方程的数值求解,是实现基于物理模型的织物仿真的技术关键。针对具有高精度要求的织物动态仿真,提出利用显式Runge-Kutta高阶方法解决仿真过程中运动微分方程的求解问题,并采用内嵌的Cash-Karp计算格式,从自动调整步长和控制精度两方面优化算法。实验证明,基于内嵌的Runge-Kutta自适应法精度高,易实现,具备良好的稳定性,并可直接获取误差,自动调整步长确保精度,使系统的求解效率得以进一步提高。

The crucial problem in cloth simulation based on physicalmodel is how to solve the motion differential equations. An explicit high-order Runge-Kutta method to solve the motion differential equations for accurate cloth simulation is presented. By use of embedded cash-karp runge-kutta formula, the algorithm is further optimized on the side of auto-adjustment for step size and error control. With high precision, less work and good stability, the method can also get the error directly and auto-adjustment for step size without loss of precision. The experiment demonstrates simulation efficiency of the whole system is improved.