基于肌肉的人体运动仿真研究

Study of simulations of virtual human movement based on muscle

针对当前国内缺少满足牛顿运动定律的人体运动仿真问题,提出了基于肌肉的人体运动仿真方法。该方法首先建立了基于解剖学的人体骨骼肌肉模型,然后进行运动捕获数据与人体模型的匹配,其次构建了主要关节点的局部坐标系,接着运用多刚体运动学与动力学进行运动学及动力学参数的计算,之后施加空间力学平衡约束得到了有效的关节运动约束机制。实验结果表明,利用该算法实现的人体运动仿真真实、自然、逼真,证明了算法的有效性。

Aiming at the questions that the current domestic is lacking in the human motion simulations which obey Newton＇ s laws of motion, a method of simulations of human movement based on muscle is proposed. The method firstly establishes a musculoskel- etal model based on anatomy; then, processes the motion capture data to match the musculoskeletal model; next, calculates the ki- nematic and kinetic parameters by multi-body kinematics and dynamics and applies the mechanical balance of space constraints in order to obtain the effective joint motion constraint mechanism. Experimental results show that the simulation of human motion is real, natural and realistic by using the proposed algorithm, and demonstrates the effectiveness of the algorithm.