基于非线性规划的人体上肢运动研究

Research on Human Upper Limb Locomotion Based on Nonlinear Programming

该文对非线性规划技术在人体上肢运动中的应用进行了研究,对解逆运动学问题的几种数值方法进行了分析和比较。然后根据这些方法的比较结果,该文选用了BFGS变度量法来求解与逆运动学问题等价的最小化问题,这种方法有良好的数值稳定性且对奇异矩阵不敏感,计算代价不高,程序设计也比较简单。最后,针对上肢的小范围运动设计了仿真系统,并和其他几种解逆运动学问题的方法在计算时间上进行了比较,结果表明了该方法的有效性和可行性。

In this paper, the nonlinear programming techniques are studied for the application to human upper limb locomotion, and several numerical methods for solving the inverse kinematics problem are analyzed and compared. Then, the BFGS variable metric method is selected to solve a minimization problem which is equivalent to inverse kinematics problem. This method is numerically stable and not sensitive to singular matrix, simple in program and low in computational cost. At last, a simulation system is designed for the small range motion of upper limb, the computational time for the several methods is compared and the result proves the validity and feasibility of this method.