Adaptive 3D Visual Servoing of a Scara Robot Manipulator with Unknown Dynamic and Vision System Parameters
Round 1
Reviewer 1 Report
The authors of the paper presented some simulation result without discussing extensivily the obtain result. The analysis of the proposed method interms of pros and cons is missing as well in the paper.
Scara is a very well know robot and so many works studied different aspects and control methods so these work must be included and sufficiently discussed in the paper.
the equations for example Kinematics and dynamic, of the robot can be found in any standard text book so can't be considered as the contribution. They can be removed, and the discussion section must be added to deeply discuss the result, pros and cons of the proposed method as well as a clear direction to the future work.
Author Response
Response to Reviewer 1 Comments
Point 1: The authors of the paper presented some simulation result without discussing extensively the obtain result. The analysis of the proposed method in terms of pros and cons is missing as well in the paper.
Scara is a very well know robot and so many works studied different aspects and control methods so these work must be included and sufficiently discussed in the paper.
the equations for example Kinematics and dynamic, of the robot can be found in any standard text book so can't be considered as the contribution. They can be removed, and the discussion section must be added to deeply discuss the result, pros and cons of the proposed method as well as a clear direction to the future work.
Response 1: We appreciate the comments and suggestions made to our article in order to improve our contribution in this field of research.
The suggested “Discussion” section was added to the article, where the simulation results obtained and the proposed method are discussed.
The robot model section was kept in order to provide details of the identified model used in the simulations.
In order to improve the “Introduction” section, it was extended to include updated works and then, in the “Discussion” section, the proposed method was compared against the existing ones.
Reviewer 2 Report
- Authors make asseverations without any analytical support
- Paper main contribution is not clear
- Authors should explain detailed all the proposed methodology in order that readers can replicate proposed results
- It would be interesting if authors can include experimental results or at least explain how the proposed methodology can be implemented for real-life problems
- Authors should justify their improvements against previous results
- Stability analysis for closed loop systems is missing
- Authors should include comparative results against existing ones
- Figure labels are in Spanish, please review
- Authors should justify the selection of the proposed robot
- Authors should justify the selected simulation profiles
- References should be updated and improved
- Figures 7 and 9, present high frequency dynamics, please explain
Author Response
Response to Reviewer 2 Comments
We appreciate the comments and suggestions made to our article in order to improve our contribution in this field of research. Here are the corrections made in response to your comments and suggestions:
Point 1: Authors make asseverations without any analytical support:
Response 1: We have revised the paper in order to make sure the asseverations and conclusions are theoretically supported by an appropriate analysis. Main theoretically conclusions are developed in Sections 4 and 5.
Point 2: Paper main contribution is not clear:
Response 2: The “Discussion” section was added to the article, where the main contribution of the article is now explained.
Point 3: Authors should explain detailed all the proposed methodology in order that readers can replicate proposed results:
Response 3: Control algorithms are given in Sections 4 and 5, and all parameters for the models as well as for the controllers are given in Section 6. We consider that this information allows the readers to replicate the results shown in the paper.
Point 4: It would be interesting if authors can include experimental results or at least explain how the proposed methodology can be implemented for real-life problems:
Response 4: The “Vision System” section (Section 3) was expanded to provide details for an experimental platform that could be used in a real-life problem. Also, all parameters values for the used models and controllers are provided in Section 6, such that the simulations can be replicated or possible experimentations could be implemented.
.
Point 5: Authors should justify their improvements against previous results:
Response 5: In the “Discussion” section the improvements of this work against a previous work were discussed, which is mainly its extension to the case of 3-D movement of the robot and the improvement to follow trajectories where the depth of the target varies with time.
Point 6: Stability analysis for closed loop systems is missing:
Response 6: The proposed control system is mainly based on two controllers, a kinematic controller and a dynamic compensator. The analysis or each loop and the interconnection is described in Sections 4 and 5, which allows to conclude about the convergence of the errors taking into account the whole closed-loop system.
Point 7: Authors should include comparative results against existing ones:
Response 7: The “Introduction” section was extended to include updated works and then in the “Discussion” section the properties and advantages of the proposed method were compared against the existing ones.
Point 8: Figure labels are in Spanish, please review:
Response 8: The labels in Spanish of Figures 5, 7, 8 and 9 of the “Simulation” section have been corrected.
Point 9: Authors should justify the selection of the proposed robot:
Response 9: In the extension of Section 3, the use of the proposed SCARA robot was justified, mainly because it is a common robotic structure referred to in many papers, and because it is available as open-software at our lab.
Point 10: Authors should justify the selected simulation profiles:
Response 10: In the “Discussion” section the selection of the selected simulation profile was justified mainly to show the performance of the controller without speed measurement under the most demanding operating conditions.
Point 11: References should be updated and improved:
Response 11: The “Introduction” section has been extended and improved to include updated works.
Point 12: Figures 7 and 9, present high frequency dynamics, please explain:
Response 12: The high frequency dynamics shown in these figures is mainly due to the use of the kinematic controller to try to control a robot (modeled with its complete dynamic model including the dynamics of its actuators) with a different structure for which the kinematic controller was designed. It was now clearly explained in the "Discussion" section.
Round 2
Reviewer 2 Report
Most of my previous comments, have been answered in this new version of the paper. However, I disagree with authors answer to my following comment:
Point 6: Stability analysis for closed loop systems is missing:
Response 6: The proposed control system is mainly based on two controllers, a kinematic controller and a dynamic compensator. The analysis or each loop and the interconnection is described in Sections 4 and 5, which allows to conclude about the convergence of the errors taking into account the whole closed-loop system.
I cannot accept this answer due to, for nonlinear system it is wrong separate the stability analysis for each part of the system. For nonlinear systems it is mandatory to do a stability analysis for the entire system.
Author Response
Response to Reviewer 2 Comments
Point 1: Most of my previous comments, have been answered in this new version of the paper. However, I disagree with authors answer to my following comment:
Point 6: Stability analysis for closed loop systems is missing:
Response 6: The proposed control system is mainly based on two controllers, a kinematic controller and a dynamic compensator. The analysis or each loop and the interconnection is described in Sections 4 and 5, which allows to conclude about the convergence of the errors taking into account the whole closed-loop system.
I cannot accept this answer due to, for nonlinear system it is wrong separate the stability analysis for each part of the system. For nonlinear systems it is mandatory to do a stability analysis for the entire system."
Thanks for your comment regarding the stability analysis and the opportunity to make a more clear explanation of our analysis. Our reasoning in getting the stability conclusion of the whole control system is the following. The control structure proposed in Section 5 of the paper is composed of two controllers connected in cascade: The kinematic controller (as described in Section 4), providing with velocity references to the second controller which is the dynamic compensator. In our reasoning, we first prove the stability conditions for the kinematic controller, obtaining the ultimately bounded error conclusion with the bound given by (29). In view of this conclusion, it becomes necessary to compensate the dynamics of the robot to improve the performance of the control. Therefore, we introduce a dynamic compensator in cascade, which is proved to asymptotically drive the velocity error to zero. In view of this result, resourcing to the backstepping reasoning, and similar to what it has been made in some of our and other authors´ previous published papers [1], [2], [3], [4], we apply this result (that is ū converging to zero) to equation (29), and conclude that it is asymptotically restored the bounded error conditions of equation (26) and the observations given there. In summary, the dynamic controller improves the stability bounded conditions even with uncertain robot dynamics. We have made a change in the last paragraph of Section 5, taken away the convergence to zero conclusion and given instead the final bounded condition, which is the stability condition for the whole control system (with kinematic controller in cascade with the dynamic compensator).
- Martins, F.N.; Brandão, A.S. Motion Control and Velocity-Based Dynamic Compensation for Mobile Robots. In Applications of Mobile Robots; Hurtado, E.G., Ed.; IntechOpen: Rijeka, 2019; chapter 2. doi:10.5772/intechopen.79397.
- Ortiz, J.S.; Navarro, G.P.; Andaluz, V.H.; Recalde, L.F. Three–Dimensional Unified Motion Control of a Robotic Standing Wheelchair for Rehabilitation Purposes. Sensors 2021, 21. doi:10.3390/s21093057.
- Andaluz, V.; Roberti, F.; Toibero, J.M.; Carelli, R. Adaptive unified motion control of mobile manipulators. Control Engineering Practice 2012, 20, 1337–1352. doi:https://doi.org/10.1016/j.conengprac.2012.07.008.
- Martins, F.N.; Sarcinelli-Filho, M.; Carelli, R. A Velocity–Based Dynamic Model and Its Properties for Differential Drive Mobile Robots. J. Intell. Robotics Syst. 2017, 85, 277–292. doi:10.1007/s10846-016-0381-9.
