Takagi Atsushi, Li Yanan, Burdet Etienne
IEEE Trans Haptics. 2021 Apr-Jun;14(2):421-431. doi: 10.1109/TOH.2020.3039725. Epub 2021 Jun 17.
Recent studies on the physical interaction between humans have revealed their ability to read the partner's motion plan and use it to improve one's own control. Inspired by these results, we develop an intention assimilation controller (IAC) that enables a contact robot to estimate the human's virtual target from the interaction force, and combine it with its own target to plan motion. While the virtual target depends on the control gains assumed for the human, we show that this does not affect the stability of the human-robot system, and our novel scheme covers a continuum of interaction behaviours from cooperation to competition. Simulations and experiments illustrate how the IAC can assist the human or compete with them to prevent collisions. In this article, we demonstrate the IAC's advantages over related methods, such as faster convergence to a target, guidance with less force, safer obstacle avoidance and a wider range of interaction behaviours.
最近关于人类之间物理交互的研究揭示了他们读取伙伴运动计划并利用其改善自身控制的能力。受这些结果的启发，我们开发了一种意图同化控制器（IAC），它能使接触式机器人从交互力中估计人类的虚拟目标，并将其与自身目标相结合来规划运动。虽然虚拟目标取决于为人类假设的控制增益，但我们表明这并不影响人机系统的稳定性，并且我们的新方案涵盖了从合作到竞争的连续交互行为。仿真和实验说明了IAC如何协助人类或与他们竞争以防止碰撞。在本文中，我们展示了IAC相对于相关方法的优势，例如更快地收敛到目标、以更小的力进行引导、更安全的避障以及更广泛的交互行为。
