A methodology for identification of uncertain LFR model of the human operator for telemanipulation with force-feedback

In this paper, a method for determining an uncertain model is proposed and applied to the identification of the human operator in a telemanipulation scheme. The method consists of two steps: first, several LTI models are identified around different configurations; second, an uncertain linear fractional representation is derived so that it encompasses all the LTI models. As an originality, the approach do not require interpolation of the LTI models and the uncertain parameters do not need to have physical meaning. The method is then applied to identify an original uncertain model of the human operator from experiments. The model is then used to analyse the robustness of a teleoperation setup.