Tunnel-shaped potential force fields for improved hand-guiding of robotic arms

Abstract

In this paper we introduce a new approach for improving hand-guiding of robotic arms. We present the tunnelshaped potential force field (TSPFF) which guides the user along a reference trajectory in order to improve the user comfort of handling robots. The reference trajectory is computed using the learning data of experienced users. The TSPFF is realized by an impedance control of the robot. The main focus of this paper is to find a robust parametrization of the TSPFF. For this purpose the stability of user-robot interaction is analyzed by means of the Lyapunov stability theory. The results are complemented by simulation studies in which the influence of the parametrization on control performance and user comfort is analyzed. The resulting rules for parametrization are validated by experimental studies with a robot with six degrees of freedom.

Stability of Nonlinear Time-Delay Systems Describing Human–Robot Interaction

“…For this reason, some authors use robust human models [3]. However, assuming the mass of the human arm is significantly smaller in relation to the system mass, it is appropriate to neglect the mass of the human arm [13], [14] or additionally the damping of the arm (for a significantly smaller human damping) [2], [15], [16]. Other human models describe the human behavior as a PD-controller [4]- [6], which could be interpreted as an active spring-damper feedback.…”

“…A comparable situation appears using a combination of impedance control and a virtual environment. These virtual environments are used in different applications such as rehabilitation [20]- [22] and industrial tasks [15], [16], [23] in the form of virtual springs, virtual walls or potential force fields. It is conceivable that a virtual environment provokes an active reaction of the human operator.…”

“…It is conceivable that a virtual environment provokes an active reaction of the human operator. Therefore, Müller et al [15], [16] chose an active behavior in the form of a positive feedback of their human model.…”

“…The simulation study of [2] is used to analyze the influences of the model parameters on the stability of human-robot interaction. Müller et al [15], [16] analyzed their system with the linear matrix inequality (LMI) approach of [24] and [25, p. 191] resulting in a numerical stability bound. Duchaine and Gosselin [13] derived analytical stability bounds in the case of a neglected response time.…”

“…Through the integration of a potential force field, the system describing pHRI could become nonlinear. A tunnel-shaped potential force field is presented in [15] and [16], in which the force is polynomial increasing depending on the distance from the zero point. This force field is used to guide an operator to a target while performing an industrial task.…”