Talk:Robot calibration

Calibration vs. Compensation

This article is dedicated to robot kinematic accruary enhancement. The mentioned method to do this is the so called "parameter identification".

This defines robot calibration in a way that methods like mapping of the working space ("volumetric compensation" for example) or of the configuration space ( Patent WO/2013/014056 for example) are excluded.

Is it intended to define "robot calibration" in a way that fault-pose COMPENSATION methods are excluded in general?

Christian — Preceding unsigned comment added by 79.141.173.22 (talk) 06:12, 3 March 2013 (UTC)[reply]

Volumetric compensation

I think "volumetric compensation" maps the working space at the end, but there are better and more direct examples for such a mapping.

A mapping of the working space is a mapping that maps a number of faulty poses to corrected poses, that are defined by artefakts for example. The rest of the working space is compensated by means of interpolation then. Compensation via - direct- working space mapping has a lot of disadvantages. — Preceding unsigned comment added by 79.227.148.117 (talk) 17:29, 7 March 2013 (UTC)[reply]

Calibration suggestions

Hello,

I think that robot calibration is not the same as parameter identifikation. Lukas Beyer, who is cited in the literature section, writes in his dissertation "Genauigkeitssteigerung von Industrierobotern":

"Unter dem Begriff „Kalibrieren“ versteht man im Allgemeinen das Feststellen des Zusammenhanges zwischen Ausgangsgrößen und Eingangsgrößen (DIN 1319, [DIN95])." So, in this regard the formula "robot calibration"= "parameter identification" appears as a unneccessary narrowing.

Since I am not experienced in Wikipedia Article maintenance and I do not want to annoy anybody, I write my suggestions here in the Talk Section.

1. I would suggest the following Change:

Parametric Robot calibration is the process of determining the actual values of kinematic.....

2. And then I would like to add:

Non-Parametric calibration on the other hand are all ways of robot calibration that do not use parameter identificarion. In general and normally, this methods are based on mappings in the work space or the configuration space.

3. Then, I would like to add in the Literature section the following article:

Gottlieb , J.: Non-parametric Calibration of a Stuart Platform. In: Proceedings of 2014 Workshop on Fundamental Issues and Future Research Directions for Parallel Mechanisms and Manipulators July 7–8, 2014, Tianjin, China

Oliver --PentapodenTentakel (talk) 19:01, 5 August 2014 (UTC)[reply]

XLINKBot need attention

Please, can anybody re-establish the link that I added yesterday in the article? The link is a real enrichment of the article. regards — Preceding unsigned comment added by Gluecksklee (talk • contribs) 05:38, 10 June 2017 (UTC)[reply]

Dear Billinghurst

I addad an external worthfully link that was deleted by a wikipedia robot. I cannot discuss with machines. But now, as far as I can see, you are made of blood, bones, etc. , and parhaps feathers (?). Now you did what the robot did before. Why? Please, let me know whether you checked out the link. Simply let me know what the topic is. There's a reason for everything, but I fear not in this case.... Best regards and have a nice weekend--- — Preceding unsigned comment added by 77.234.45.159 (talk) 16:18, 16 June 2017 (UTC)[reply]

Parallel Kinematics Parametric Robot Calibraion

Jean-Pierre Merlet states in "Parallel Robots: Open Problems",

https://www-sop.inria.fr/teams/hephaistos/PDF/merlet_isrr99.pdf

the following:

"..... Even if a quite accurate estimates of

these parameters are available, a calibration may be

necessary. Although this problem has been solved for

serial robots, this is not the case for parallel robot."

and

".... Applied to parallel robot this method leads

to calibration result that are in general disastrous."

Regards — Preceding unsigned comment added by 46.223.163.210 (talk) 09:28, 12 November 2021 (UTC)[reply]

Parallel Robot Calibration

There is a new approach to calibrate parallel kinematics, that does not fall in a category metioned in this article: "Verfahren und Anordnung zur hochgenauen Kalibrierung einer Parallelkinematik", DE 10 2018 124 898 A1. — Preceding unsigned comment added by 79.247.143.97 (talk) 13:42, 9 December 2021 (UTC)[reply]

The Parametric Robot Calibration of a Stewart Platform

Pose measurements are carried out in an auxiliary coordinate system, which is required for the parametric calibration of a Stewart Platform. These poses result, for example, from measurements on the contour of the platform, or from measurements on a pose defining artifact attached to the platform. Measured coordinates on the surface of the platform or an artifact are not related to the parameters to be identified, which are also given as coordinates. Such references presuppose the existence of a well-defined, canonical coordinate system, referred to below as the reference coordinate system. Such a reference coordinate system must be able to be specified at any time in any other coordinate system, and it must be the basis of every calibration.

Incidentally, the surface of the platform is also usually not manufactured with high precision, and pose defining artifacts are mounted on the platform.

The position and orientation of each pose defining artifact is therefore not related to the joint coordinates, since a common reference coordinate system is not defined and therefore also cannot be specified with an auxiliary coordinate system.

With parametric calibration, all pose measurements are based on an ephemeral coordinate system - for example, a coordinate measuring machine - whose position can only be roughly compared with a fictitious reference coordinate system based on design drawings.

In summary, one has to state that parameter identifications with the Stewart Platform always implicitly presuppose a well-defined reference coordinate system, which, however, does not exist at all. Of course, this has the consequence that the parameter identification cannot determine the true geometry parameters of a Stewart Platform, but that the parameters of the Stewart Platform only serve as the parameters of the compensation function in order to be able to represent the results of the pose measurements using a few parameters. So you have to question the term "identification".

If one works with a calibrated Stewart Platform, then the user can only roughly determine the position of a fictitious existing reference coordinate system by inferring its position from measurements on the platform. However, if the reference coordinate system is not defined, then no precise statements can be made about the directions of movement of the Stewart Platform, because these directions are defined by the reference coordinate system. The same applies to the position of the rotary axes. StewartPlatform (talk) 16:28, 13 March 2022 (UTC)[reply]