神经网络及误差补偿在HIT-Ⅲ双足机器人步态规划中的应用

Application of Neural Networks in a HIT-Ⅲ Biped Robot's Gait Programming

运动学求解是双足机器人步态规划的基础。针对 HIT 双足机器人实体 ,利用 BP神经网络求解其规划运动角和反馈的实际输出角之间的非线性映射。计算结果表明 ,该法简便易用 ,计算精度高。为了满足机器人在线实时控制要求及进一步提高运算精度 。

The solution of kinematics is the base of gait programming of a biped robot. In this paper, the authors use back propagation network to solve the non-linear relation between programmed moving angle and the actual angle of a HIT- Ⅲ biped robot. The calculated results suggest that this way is simple and easy to use, and the precision is high. In order to meet the requirements of on-line and real-time control of robot and to enhance the precision, a way of error compensation using iterative calculation.