柔性杆子杆模型参数的遗传算法优化求解

研究了应用实数型的遗传算法解决柔性机器人子杆法建模中,子杆模型参数的优化求解问题.子杆模型参数形成高度非线性的方程组,它的求解是子杆法能否实现的关键.普通的牛顿迭代法及其变形,要求初始向量足够靠近解,一些所谓大范围收敛的算法又有方程组的Jacobian矩阵非奇异等要求,在此都难以应用.本文基于遗传算法的解法可求得模型参数的最优解,实例仿真显示了这一方法的可行性和高效率.

Using Genetic Algorithms to Solve Sub-Link-s Parameters of Flexible Link

A genetic algorithm of real number type is used to solve the equations of the sub-link-s parameters when modeling a flexible robot manipulator with the sub-links approach. It is the key of successfully using the sub-links approach that how to solve the equations of the sub-link-s parameters. However, the set of equations is high non-linear. The general solving approach is hardly fit for it because of lack of initial vector that enough close to the solution. Some approach possessed the ability of large-scale convergence is failing too because of the singularity of the Jacobian matrix of the equations. The proposed approach based on genetic algorithm can smoothly solve the equations by optimization method that replace some complicated symbolic calculation by solving simple algebra equations. Then the validity and high efficiency of the proposed approach are showed by a practical example.