Push Recovery Control of a Bipedal Robot Standing on Two Offset Planes in Double Leg Stance
Vyankatesh Ashtekar, Ashish Dutta
- 发表年份
- 2023
- 引用次数
- 2
摘要
This paper implements two push recovery control strategies for a twelve-degree-of-freedom bipedal robot subject to impact loads while standing in a double leg stance with its feet on offset horizontal planes. In the ankle strategy, the robot’s centre of mass and the torso orientation are regulated at a set point. A hip strategy is proposed where the ankle strategy fails to maintain the robot’s balance in the presence of larger disturbances. The torso performs a pitching motion about an axis defined by tipping conditions while regulating the prescribed centre of mass position. Both strategies are developed to suit stiff position-controlled actuators. The joint angle trajectories are found using position and velocity inverse kinematics and enforced using a proportional controller. Accurate knowledge of the robot balance is necessary for switching between ankle and hip strategies. An issue in defining stability margin in the sense of dynamic balance of a biped is highlighted in the context of a double leg stance on the split ground. The pseudo zero-tilting moment point is used to evaluate balance in this context, and its movement in the presence of varying loads is examined. The push recovery simulations are performed on a miniature biped robot modelled in MuJoCo, an open-source library with C API. Numerical simulations in MuJoCo analysing bipedal dynamic balance and push recovery strategies are presented. An inverse kinematics based stance planning utility for bipeds is developed in MuJoCo as a part of this work.
关键词
相关论文
Statistical Learning Theory
Yuhai Wu, Vladimir Vapnik
1999
Artificial intelligence: a modern approach
1995
Applied Nonlinear Control
Jean-Jacques Slotine, Weiping Li
1991
A new optimizer using particle swarm theory
R.C. Eberhart, James Kennedy
2002