On the use of shunt impedances versus bounded environment passivity for teleoperation systems

Abstract

This paper analyses and compares two passivitybased approaches that allow to include a-priori knowledge on the dynamic range of the human operator and/or the environment. This can lead to less conservative teleoperation systems compared to systems designed to be purely passive or absolutely stable. The first approach under investigation is a method where the absolute stability is analysed of a teleoperation system augmented with shunt impedances in series and/or parallel with the teleoperation system. It is shown that the traditional interpretation of the use of shunt impedances is not valid and a more accurate description of how to use this method is presented. The second approach under investigation is the bounded environment (operator) method. It is shown that the original idea to restrict the analysis to the so-called worst-case scenarios of a pure mass and a pure stiffness as environment can be too simplistic. Illustrative examples with mass-spring-damper systems fixed to the ground and floating objects as environments are made to demonstrate this in detail. In conclusion, this paper shows that embedding environment knowledge into the controller analysis/design is not straightforward and further research should be dedicated to determine which bounds should be used to obtain practically stable systems for different applications.

Towards collaborative drilling with a cobot using admittance controller

“…Although Willaert et al (2011) and Tosun and Patoglu (2020) showed that higher physical damping may yield in a non-passive system under some special conditions, we know that such systems are not necessarily unstable (Buerger et al, 2001). To this end, our result complements the findings by Willaert et al (2011) and Tosun and Patoglu (2020), and suggests that the upper bound of the physical damping should be considered for the stability analysis in addition to its lower bound.…”

“…Moreover, it is widely accepted that b eq contributes to the stability favorably (Colgate and Brown, 1994;Colgate and Schenkel, 1997;Colgate et al, 1993;Weir and Colgate, 2008), thus, only the lower bound of b eq = 0, is considered as the worst case in the earlier studies. However, Willaert et al (2011) and Tosun and Patoglu (2020) presented several examples where larger b eq may result in an active (i.e., non-passive) system under specific conditions, though the system is passive for b eq = 0. Instead of passivity, we directly investigated the stability in our analysis, and observed that our results do not indicate such a situation within the limits of the equivalent impedance considered.…”

“…In utilizing the bounds of human arm and/or environment impedances, it is typically assumed that stiffness and mass contributions of these impedances may adversely affect stability. However, recent research shows that a special care should also be given to the damping component (Tosun and Patoglu, 2020; Willaert et al, 2011), though it has been assumed so far that increase in damping contributes to stability favorably (Colgate and Brown, 1994; Colgate and Schenkel, 1997; Colgate et al, 1993; Weir and Colgate, 2008). In particular, it was shown that high damping may result in non-passive systems under specific conditions, so it may as well be dangerous for stability (Tosun and Patoglu, 2020; Willaert et al, 2011).…”

M-Width: Stability, noise characterization, and accuracy of rendering virtual mass

“…Representative robot and environment parameters. The device parameters were inspired by Diolaiti et al (2006), and the environment parameters by Kuchenbecker et al (2003) and Willaert et al (2011). the continuous model whose stability boundaries are similar, but not identical, to the linear hybrid and discrete model stability boundaries.…”

Bilateral Teleoperation in Cartesian Space with Time-Varying Delay

“…The classic scattering transformation or wave variables approach would lead to position drift [5] [6] and has a trade-off between stability and system performance [7] [8].…”