A Novel Adaptive and Nonlinear Electrohydraulic Active Suspension Control System with Zero Dynamic Tire Liftoff

Round 1

Reviewer 1 Report

The research topic is interesting. The reviewer's main concerns are:

a) In the methodology part, lots of formulas were from the references, such as [6], [12], [18], etc. Also, in the Introduction part, the authors mentioned that "we apply a nonlinear control filter that was developed in [6]".

This poses the problem of sufficiency of novelty as a journal publication.

b) Numerous typos and grammatical errors are found throughout the paper. Some sentences need to be rephrased. 

For example, line 64 on page 2, "However, there is no control performance results were showed when...."

c) The literature review is not sufficient. More work needs to be done. Besides, in line 57 on page 2, "several recent works [7-12]...." More details should be provided for each cited reference.

d) The presentation should be improved. For example, i) Figure and its caption should be kept within one page. ii) On page 7, description for the backstepping control is not professional. The current form is like a technical report. Also, no equation numbers are given. Most formulas in this part are not very readable. 

e) In Section 4.5 Simulation and result discussion, for Table 1 on page 13, why to choose these parameter values? Also, the quality of results figures is rather low. 

f) In the References part, the citation form is not standard. For example, Refs. [4], [9] and [11] should provide the correct journal names, instead of using "Elsevier". The author names should be clearly given, please check Ref. [15], [17], etc. 

Author Response

Thank you very much for providing these valuable comments and suggestions.

Answers

a)

We use those references in the methodology to prove the formulas’ sources that many MDPI articles were written using this style, such as “An Active Vehicle Suspension Control Approach with Electromagnetic and Hydraulic Actuators.” On the other hand, we remove some references in the methodology that we mentioned in the introduction.

We redesign the nonlinear control filter from two regimes introduced in [11], to three regimes in this study to address our new modeling of a dynamic landing tire system. The novelty in this study is explained as follows: lines (111-127), (150-153), (156-160), (161-194).

b) 

Thank you so much, English mistakes are corrected.

c)

We add more literature review and provide more details for works [7-12], as explained in lines (31-110).

d)

we improve presentation of all paper sections.

     i) All figures and their captions are adjusted.

      ii) The backstepping control description is rewritten professionally as in lines          (209-217). Also, the radial basis function neural networks RBFNN is      described in lines (219-243).

       iii) we number the paper’s formulas.

e)

      i) In Section 4, we choose these parameters based on the literature review setup and the control sensitivity.

     ii) we modify passive suspension simulation in Figures 4, 5 and 6 to find the percentage improvements between filtered and unfiltered active suspensions and also the passive suspension. We also improve results in Figures 7, 8, 10 by applying bumpy and pothole road design instead of bumpy road design to clearly address the tire liftoff. In section 4, all result discussion is qualified by introducing percentage improvement in control systems and passive suspension.

f)

We used EndNote, which may have mistakes. In this revision, we fix all references.

Reviewer 2 Report

This paper is well written and very interesting. Before the publication, some issues need to be addressed.

1. The novelty and contribution of this work need to be further clarified.

2. The literature review could be improved by discussing some related works in active controls and their applications in automotive systems, for example: High-precision hydraulic pressure control based on linear pressure-drop modulation in valve critical equilibrium statel driving-style-based co-design optimization of an automated electric vehicle: a cyber-physical system approach; Dynamic State Estimation for the Advanced Brake System of Electric Vehicles by using Deep Recurrent Neural Networks.

3. Clearly clarify the pros and cons of your proposal.

4. Some future work could be included in the last section.

Author Response

Thank you so much for reviewing these valuable comments and suggestions.

Answers

1)

We clarify the novelty and contribution, which explained clearly in lines (111-127). Also, we define the novelty methodology as in lines (156-170).

2)

We extend the literature review by introducing extra references and explain more details for some of them as in lines (31-110). Also, we present those references that you suggested as in lines (44-56).

3)

In lines (25-27), we clarify the control effectiveness.

4)

In conclusion line 541, we suggest a future work that we will study.

Reviewer 3 Report

P0: 

Introduction can be extended with some standards of tests (testing impulses) for suspension system in cars and what requirements are defined for comfortable status.

P1:

Figure 6. filter reaction time is usually measured not from graph time zero, but from start of road displacement, please rewrite correct values. Same problem with all next figures.

P2:

RBFNN has one hidden layer is fixed according description, but it is not defined how many nodes was used.

P3:

Test road impulse is very simple, will be better include two oposite road impulses (or same) in short time for example inside 1 second to test stability of this new algorithm approach.

P4:

Conclusion doesn’t contain any real acquired values: sentence "better suspension" is not enough for research article. Please extend conclusion with some evaluation for example percentage improvement to passive system.

Author Response

Thank you so much for reviewing these valuable comments and suggestions.

Answers

A0:

we extend the literature review by introducing extra references, explain more details for some previous works as written in lines (31-110). The extending references that include testing impulses are written in lines (57-66). All comfortable states are explained in the first paragraph of introduction lines (23-43) and (57-61).

A1:

Figure 6. filter reaction time is usually measured not from graph time zero, but from start of road displacement, please rewrite correct values. Same problem with all next figures.

A1We modify passive suspension simulation in Figures 4, 5 and 6 to find the percentage of improvements between filtered and unfiltered active suspensions and also the passive suspension. We also improve results in Figures 7, 8, 10 by applying bumpy and pothole road design instead of bumpy road design to clearly address the tire liftoff. In section 4, all result discussion is qualified by introducing percentage improvement in control systems and passive suspension.

A2:

In lines (219-242), we extend RBFNN description and explain how to tune the RBFNN’s parameters.

A3:

In Case II lines (448-452), we extend introducing bumpy and pothole impulse road designs to address the car-road stability. Because of the high road frequency, Case II can generate tire liftoff phenomenon, as shown in Figures 7, 8, 10.

A4:

We add percentage improvements in the proposed control system NAC for each case comparing with unfiltered active and passive suspensions, as explained in lines (534-540).

Round 2

Reviewer 1 Report

The manuscript can be accepted in its current form.

Author Response

Minor English grammar mistakes are deleted or adjusted as show in highlighted marks.

Thank you