Robust Safe Motion Control for Compliantly Actuated Robots via Disturbance Observers
Abstract
Compliant actuators are commonly utilized in physical interactions between humans and robots, and it is of great significance to focus on safety control issues. This article introduces a robust safe motion control (RSMC) framework that employs control barrier functions (CBF) for robots driven by compliant actuators. Compliantly actuated robots are commonly subject to both matched and mismatched time-varying disturbances, including external environmental disturbances, imprecise link parameters, and unknown loads. These factors and their higher order derivatives can have adverse effects on CBF-based safety control, resulting in safety violations and degraded control performance. To ensure robustness of safety against disturbances, a new disturbance estimates-based high-order control barrier function (DE-HoCBF) is constructed by fully utilizing both the disturbance estimates and the upper bound of the estimation error. On the basis of the constructed DE-HoCBF, the RSMC law is established by solving a quadratic programming problem. Compared with other worst-case-based robust CBF methods, the RSMC method proposed in this article achieves a better tradeoff between robustness and safety. Experimental results are provided to validate the effectiveness of the proposed method.
