A Modeling and Control Algorithm for a Commercial Vehicle Electronic Brake System Based on Vertical Load Estimation
Round 1
Reviewer 1 Report
In this paper, the authors adopt an Unscented Particle Filter method to estimate the vertical load during the vehicle traveling and to achieve an accurate control in the pneumatic pipeline. The approach was corroborated via the Amesim software, whose a deep analysis was dedicated. In the following, all the comments to reply.
1. The literature overview needs an improvement with some recent results. See for example https://doi.org/10.1504/IJVD.2015.068134, https://doi.org/10.1016/j.jfranklin.2022.07.015, DOI: 10.1109/ACCESS.2023.3262023, DOI: 10.1109/TIM.2022.3180407 and the references therein.
2. The authors should provide all unit measures for the model parameters (e.g. the air density, the contact area, etc.). They could be added to Table 1.
3. I disagree in considering the Particle Filter algorithm as an extension of the Kalman Filter one. The theory behind is different.
4. The second contribution (lines 82-85) needs a reformulation, avoiding the abuse of the "control module". Moreover, "commercial vehicle EBS control algorithm is proposed" is unclear information: which kind of control has been designed? A clearer and more exhaustive explanation is crucial in a contributions list.
5. Are not completely clear the differences between the PF and the UPF approaches.
6. The reasons behind the choice of The UPF solution deserve a clear explanation and verification, for example in the simulative section the authors could provide a comparative analysis with respect to some classical methods, as the Kalman-filter based approaches, specifically from the computational burden point of view.
7. It is interesting the state variables vector given in eq. (5) and, specifically, the presence of the mass m, which I think can be considered as a constant value. This means that the UPF approach exploited serves also as a parameters observer? Moreover, is it possible to consider the mass with the constant nominal value? What is its time-dependent model? I think that its time-varying changes and uncertainties are not observable on the long time and could be neglected.
8. Give a clear definition of the measure model in (13) and justify the choice of the matrix H.
9. In the simulation verification, the authors should provide a correct space between the measurement value and the unit measurements.
10. It seems that the paper's aim is directed toward the estimation task, with verification via the Amesim software. Accordingly, in the Introduction section, reviewing the state-of-art, it is useless to focus the attention on the control solutions, since this is another task beyond the scope of this work because the authors did not propose a proper control solution.
11. Some typos are present. For example “EBS integrates components such as sensors, control units, and actuators” is better than “EBS integrates components such as sensor, control unit, and actuator”; “at the current and the last sampling time” than “at the current sampling time and the last sampling time”; "designed a new EBS system based on an EBS" is a sentence that needs a clearer explanation; “structured modeling of the EBS” than “structured modeling of EBS”, “as follows: the vertical” than “as follows: The vertical; at lines 96-100 the list should contain a comma than a point; it is not correct to introduce a formula with a point (see relations (5)-(6)-(7)-(11)-(12) etc.); A careful rereading is recommended.
Extensive editing of English language required
Author Response
Please see the attachment.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The authors present a control strategy for a pneumatic pipeline relying on the vertical load estimator based on Unscented Particle Filtering (UPF). The control algorithm of the valve was studied, and the algorithm was tested using hardware in the loop experiment, demonstrating interesting advances in improving braking performance. The paper has to be considerably improved in terms of the overall formulation of modelling hypotheses and parameters validation, but the research potential outcomes are very promising.
It is not clear which is the reason references [21-24] are reported since they do not seem to be connected with the topic under investigation. Furthermore, the introduction section must be definitely improved mainly addressing the control approaches from the theoretical point of view, specifying their limitations or application domain. It becomes how the described approach can be integrated with the currently adopted estimators and controllers and and how this kind of approach can be generalized? Additionally, to enhance the overall discussion, I would specifically suggest to include the control strategies referring the tyres’ management in contact with completely different terrains (different stiffnesses, friction levels, thermodynamic conditions, etc.) and which characteristics could be modified significantly as function of its thermal and wear state both during a singular maneuver and through the entire lifecycle. Furthermore, another very important aspect regards the pollution due to particles of worn rubber and brakes. I would also specify that the current study does not include these specific topics and they represent further important aspects to be considered within the design of the future control strategy, considering the entire vehicle and all its subcomponents. Among the studies to discuss, I would definitely discuss the following ones:
- [1] Guo, Jinghua, et al. "Safe and energy-efficient car-following control strategy for intelligent electric vehicles considering regenerative braking." IEEE Transactions on Intelligent Transportation Systems 23.7 (2021): 7070-7081.
- [2] Santini, Stefania, et al. "On-board road friction estimation technique for autonomous driving vehicle-following maneuvers." Applied sciences 11.5 (2021): 2197.
- [3] Hellgren, Jonas, and Erik Jonasson. "Maximisation of brake energy regeneration in a hybrid electric parallel car." International Journal of Electric and Hybrid Vehicles 1.1 (2007): 95-121.
- [4] Mosconi, Lorenzo, et al. "Adaptive vehicle dynamics state estimator for onboard automotive applications and performance analysis." Vehicle System Dynamics (2022): 1-25.
Is the load effect on friction included within the tire modelling of the vehicle? An additional sensitivity on this effect (within the estimation and control logic) will be interesting to address in this study.
It is evident that there is a not negligible delay between the estimated and true value of the vertical load. Since, braking maneuvers are generally very fast (especially for heavy vehicles), I would suggest the authors to detail the application domains of their findings (types of terrains, vehicle, maneuvers, etc.).
Section 4.1 should be detailed better. A graphic scheme could be adopted to better illustrate the logic and the relative thresholds. It is not clear how the deceleration of the vehicle is governed, maximizing the tire friction ellipse: is it something relying on an ABS system or additional information concerning the wheels’ angular velocity?
All the scenarios presented are affected by the modelling hypotheses which have not been presented and discussed. For this reason, at the current stage, the results section can not be discussed properly. I would advice the authors to put in evidence the capability of the presented approach to be extended in a generalized one for any vehicle and scenario of use.
In the current form, I would suggest the authors solve the main issues of the hypotheses definition and formulation validation before resubmitting the work.
Some typos across the texts have to be fixed.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
No further questions arise. The paper can be accepted now. According to my previous question, number 9, for the correct space between between measured values and unit measurements I meant that, sometimes, in the text (not in the figures), the authors indicated 50km/h instead to 50 km/h etc., and so on for other quantities.
Reviewer 2 Report
The revised manuscript have addressed the main issues detailing and clarifying the limitations, modelling hypotheses and future development applications.
