基于虚拟支点倒立摆模型的双足机器人质心轨迹规划

针对传统的双足机器人在摆动相机械能起伏较大,其能耗远高于人类的步行模式的问题,提出了一种基于虚拟支点倒立摆被动特性的双足机器人步态规划方法,在保证摆动阶段机械能近似守恒的前提下,同时保证压力中心点能够从脚跟移动到前脚掌,与人类步行的地面反力特性接近;为解决倒立摆微分方程没有初等解的问题,提出用线性倒立摆方程估算摆动角度.数值仿真结果表明,该方法在与人接近的步行参数下获得的运动模式其效率远高于传统方法.

COM Trajectory Planning Method for Biped Robot Based on Virtual Pivot Inverted Pendulum

There are many traditional gait planning methods based on the linear inverted pendulum or table-cart model,by which the mechanical energy of robot during the swinging phase is fluctuant and leads to inefficient gait compared to human,to solve this we introduce a new gait planning method based on an inverted pendulum with a virtual pivot,by which the mechanical energy of robot is approximately conservative and the ZMP is traveling form the heel to toe,like the ground reaction force mechanism of bipedalism; To deal with the difficulties of solving the differential equation of inverted pendulums,we introduce an estimation method.A numerical simulation shows more energy-efficiency compared to the traditional method.