High dynamic balancing

Despite the standing stability criterion for robots with co-planar contacts has been established for decades, i.e., restricting the ZMP within the support polygon, it is not clear how to define the standing stability margin when the robot executes high dynamic or impact motion, see the impact-experiment by the organizer Wang et al. [1].

Extract the relatively simple dynamics of balancing from the potentially complex complete dynamics of the whole robot; and we can design balance controllers that exhibit higher levels of speed, accuracy and robustness than previously reported, on top of the recent achievement [2], our invited speaker Roy Featherstone will share his latest opinion.

Momentum-based control design has demonstrated convincing results for humanoid balancing, such as stabilizing with partial footholds by Wiedebach et al. [3]. Despite the achievements, the momentum-based control design is an active research field with open questions such as:

  • What is the best control design to decouple the angular and linear momentum?
  • For robots performing in-air motions, how to represent the momentum with respect to the earth, to its COM, and to its floating-base?

Our invited speaker Dragomir Nenchev will share his most recent answer to these questions.

[1] Yuquan Wang, Dehio Niels, Arnaud Tanguy, and Abderrahmane Kheddar. Impact-aware task-space quadratic-programming control. 2020. URL https://arxiv.org/pdf/2006.01987.pdf.

[2] Josephus JM Driessen, Roy Featherstone, and Antonios E Gkikakis. An actuator design criterion to maximize physical balance recovery. In 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 2829–2836. IEEE, 2018.

[3] Georg Wiedebach, Sylvain Bertrand, Tingfan Wu, Luca Fiorio, Stephen McCrory, Robert Griffin, Francesco Nori, and Jerry Pratt. Walking on partial footholds including line contacts with the humanoid robot atlas. In IEEE-RAS International Conference on Humanoid Robots, pages 1312–1319, 2016.