在采用线性耦合振荡器模型的仿人机器人步态规划中,振荡器部分参数依靠手动调整难以获得能稳定行走的值,故提出了一种优化振荡器参数的算法以改善整个步态规划方法.将机器人的步行速度、行走过程中的质心振动幅度和ZMP(zero moment point)曲线的回绕率组合起来作为优化目标,定义并选定ZMP稳定裕度为约束条件,采用带约束的非线性优化方法作为求解工具,在部分参数固定的情况下,计算得到线性耦合振荡器部分参数的最优值.仿真结果表明,优化方法选择的参数能在保证较大ZMP稳定裕度的情况下实现机器人的快速步行.在DARwIn-OP仿人机器人开放平台上进行的步态规划实验结果表明,优化得到的线性耦合振荡器参数能保证仿人机器人以较大速度稳定行走.
The solution of tension distributions is infinite for cable-driven parallel manipulators（CDPMs） with redundant cables. A rapid optimization method for determining the optimal tension distribution is presented. The new optimization method is primarily based on the geometry properties of a polyhedron and convex analysis. The computational efficiency of the optimization method is improved by the designed projection algorithm, and a fast algorithm is proposed to determine which two of the lines are intersected at the optimal point. Moreover, a method for avoiding the operating point on the lower tension limit is developed. Simulation experiments are implemented on a six degree-of-freedom（6-DOF） CDPM with eight cables, and the results indicate that the new method is one order of magnitude faster than the standard simplex method. The optimal distribution of tension distribution is thus rapidly established on real-time by the proposed method.
