Optimal Actuator Placement for Real-Time Hybrid Model Testing Using Cable-Driven Parallel Robots

Round 1

Reviewer 1 Report

The results presented in the paper concerns a specific area of applications, moreover the paper is well written and interesting. The approach proposed is very well posed and it has been appreciated both the theoretical aspects and the technical contributions. In some sense this type of studies could be included both as regards the models and the approach to the problems that could also be faced following the guidelines of imperfect systems. Therefore I suggest to include in the paper the following reference also to make extend the references and to interest a wider class of readers:

Control of imperfect dynamical systems Autori Maide Bucolo, Arturo Buscarino, Carlo Famoso, Luigi Fortuna, Mattia Frasca Data pubblicazione 2019/12/1 Pubblicazione Nonlinear Dynamics Volume 98 Numero 4 Pagine 2989-2999 Editore Springer Netherlands  

Author Response

We thank you for this positive feedback, and for providing the suggested reference. After consideration and given that our introduction already is quite long (with a substantial amount of references), we have chosen not to include this reference in the updated manuscript.  

Reviewer 2 Report

This paper deals with an optimal actuator placement for real-time hybrid model testing using cable-driven parallel robots. I would like to point out following as:

 

1. In introduction, I think the description of object of this paper is ambiguous even though authors describe object in line 79 through 83.
2. In table 1, authors describe the differences between CDPR in typical applications and CDPR in ReaTHM testing, however, authors should add differentiation compare with other paper.
3. Are there any differentiation between proposed method and CPS (cyber physical system)? I hope add explain relationship between CPS and proposed method in introduction.
4. Please add entire block diagram, to easily understand this paper.
5. I think this paper requires the kinematic equation and dynamic equation for cable-driven parallel robots to judge stability for cable-driven parallel. Did you check the stability?
6. Please add result by using table in section 4.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

To meet some peculiarities of real-time hybrid model (ReaTHM) testing, this paper presents a procedure for optimal actuator placement mainly based on a performance measure. The latter leverage the previous work of the author in [30] and accounts for ReaTHM testing peculiarities.

The paper well written and the work done explained.

Experiments on the prototype for ReaTHM testing obtained with the proposed procedure should be presented.

Moreover, the following comments should be taken into account.

Section 1:

About the part of the introduction dealing with placement of actuators, several relavant previous wrok on CDPR geometric synthesis and optimization are overlooked, e.g.:

K. Azizian and P. Cardou, The dimensional synthesis of planar parallel
cable-driven mechanisms through convex relaxations, ASME Journal of
Mechanisms and Robotics, 2012.

M. Gouttefarde, J. F. Collard, N. Riehl, C. Baradat, Geometry Selection of a Redundantly Actuated Cable-Suspended Parallel Robot, IEEE Transactions on Robotics, 2015.

J. T. Bryson, X. Jin, and S. K. Agrawal, Optimal design of cabledriven
manipulators using particle swarm optimization, ASME Journal
of Mechanisms and Robotics, 2016.

Section 2:

You should not use a new name, "the load feasibility requirement" for something which has already been discussed several times in previous works and which is more commonly known as "required/desired wrench set".

Moreover, this section misses to highlight that the set of feasible loads is a zonotope (pointed out in [45], S. Bouchard et al, On the ability of a cable-driven robot to generate a prescribed set of wrenches, ASME Journal
 of Mechanisms and Robotics, 2010) which can be determined efficiently by means of the procedure proved in M. Gouttefarde and S. Krut, Characterization of parallel manipulator available wrench set facets, Advances in Robot Kinematics, Springer, 2010. This determination yields the description of the set of feasible loads a linear inequality system which allows to straightforwardly test if a W-req such as the one defined in (6) can be tested.

Section 3:

It is written in the introduction that one peculiarity of using CDPR in ReaTHM is to have low stiffness. How is this reflected in the performance measure defined in eq. (8)? This probably deserves further explanations.

Besides, the sentence "f_0 is the solution f* to (4) given w_ref=0" seems to indicate that w_ref=0 is the target? If it is the case, this should be highlighted and explained in the manuscript.

Section 4:

Presentation of experiments is missing.

The fact that crossed cable configurations yields better results is well known and previous literature on this should be acknowledged.

Other possible bibliography:

R.M. Bennett, M. G. Farmer, R. L.Mohr, and W. Earl Hall Jr., Wind-tunnel technique for determining stability from cable-mounted models. Journal of Aircraft, 1978.

Minor comments:

• Page 3, "CDRP applications" --> "CDPR applications"
• f_0 is not defined. It should be clearly defined notably since it is used in (8)

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

This paper well revised according to reviewer's comment, thus I would like to decide as an "accept". 

Reviewer 3 Report

All my comments have been properly taken into account. I understand the issue associated to adding experiments to the reported work.