Hedge algebras-based admittance controller for safe natural human–robot interaction

In physical human–robot interaction (pHRI), the identification of inertia and damping matrices in the dynamic admittance model is still an open problem. Besides, the natural interaction is rarely considered in previous studies while it is crucial to obtain the effective cooperation. To this end, a fuzzy-based admittance controller is presented, in which the end-effector's velocity is adaptively adjusted via the external wrench and transmitted power without the identification of inertia and damping matrices. Besides, this fuzzy-based admittance controller also guarantees the natural cooperation between human and robot. Unfortunately, there is no formulated linkage of the fuzzy sets with the natural linguistic term semantics. As a consequence, human experts must utilize the order relationships between the terms of interest when formulating the fuzzy rule-based knowledge of the fuzzy controller. This paper presents an alternative admittance controller for pHRI based on an algebraic approach to linguistic hedges in fuzzy logic to overcome the existing shortcomings of previous admittance controllers. In addition, this paper also considers end-effector's full degree of freedom to guarantee the natural human–robot interaction. The proposed admittance controller is experimentally evaluated on a teaching task set-up using 6-DOF manipulator.