摘要
在双足机器人跨越迎面而来的动态障碍物的问题中,由于障碍物的高度,和速度是不可预测等因素,机器人的迈步步长和迈步高度决定了其能否实现成功跨越。介绍一种双足机器人步态规划方法,应用模糊Q学习算法对迈步高度进行学习,将迈步的起始点、落点和迈步高度作为特征点,利用三次样条对特征点进行插值得到摆动腿运动轨迹,最后通过摆动角间的几何关系得出各关节处摆动角的变化规律,控制机器人跨越动态障碍物。仿真结果表明,通过进行的步态规划,机器人可以成功跨越动态障碍物,并且各关节处的摆动角变化曲线平缓无畸变。
In the problem of biped robot stepping over dynamic obstacle in the sagital plan, because of the height and unpredictable velocity of obstacle, the step length and step height of robot are two key factors which influence successful stepping over. An approach of walking gait planning for biped robot has been presented. Fuzz Q - Learning algorithm has been used to learn step height. Cubic spline has been applied to interpolate the designated point such as step starting point, landing point and maximum step height to get the swingleg trajectory. The curve of pitch angle of swingleg has been obtained according to the geometrical relationship between pitch angles. The simulation resuits show that the biped robot can step over obstacle successfully, and swinglwg trajectory is smooth and there is no abrupt change for pitch angles.
出处
《计算机仿真》
CSCD
北大核心
2009年第4期218-222,共5页
Computer Simulation
关键词
双足机器人
步态规划
模糊学习算法
三次样条插值
Biped robot
Walking gait planning
Fuzzy learning algorithm
Cubic spline interpolating
