An MFD Construction Method Considering Multi-Source Data Reliability for Urban Road Networks

Round 1

Reviewer 1 Report

The paper presents an interesting problem on traffic management and control. Although the results may be useful, they are not presented well. The paper involves many typos and is not written with a good language level. The literature review should be strengthened with up-to-date subject references. Also, the structure of the paper should be improved. Therefore, I suggest either that the authors make a major revision of the paper or that the paper is rejected, and the authors can re-submit once they acceptably present the results.  

Major comments:

• The (English) language of the whole paper should be improved. Extensive proofreading by a native speaker is necessary.
• Introduction and abstract: From reading the abstract (and the Introduction), I don't understand the paper's aim, novelty, and contribution. This should be better stated. Therefore, the abstract should be rewritten, and the Introduction should also be rewritten such that it
  (a) mentions the importance of the topic,
  (b) involves a short state-of-the-art,
  (c) explains how the paper contributes to the problem,
  (d) describes the structure of the paper.

Such a 4-paragraph approach seems to be relevant for the paper. However, this should be followed by a literature review section.

• Literature review:
  (a) The paper should separately involve either a section 2 (or subsection, e.g., 1.1) dealing with the literature review.
  (b) Maybe a table involving a survey of main characteristics among literature references would be helpful. Such a table should prove and strengthen the contribution of this paper.
  (c) It is quite surprising (or rather worrying) that the literature review does not involve many recent papers (only one paper from 2020+). I am, for example, aware of several (renowned) authors and their recent papers (e.g., prof. Carlos Daganzo or Yicheng Zhang).
  (d) The authors claim that the "Macroscopic fundamental diagram provides a new way to optimize…". A new? Just by a brief look, I found the first papers involving "MFD" in the title from 2008 (by Daganzo et al., as cited). Especially this keyword (MFD) and its research progress since 2008 should be discussed in detail!
• Materials and methods:
  (a) It is necessary to involve a table (or tables) of, e.g., sets/indices, parameters, and variables used in Section 2. This could either be in Section 2 or in Appendix.
  (b) The equations are not clearly presented; for example, during reviewing eq. (5) and (6) I have realized that a scheme of an illustrative example would be helpful (to explain the various notations, parameters, variables, or principles). For example, I don't understand "..? and ? are the count of the first link and last link installed with ALPR detectors…". The scheme should help with a better understanding of the whole model.
• Results:
  (a) Section 3 is around 12 pages. In the 22 pages long paper, this seems to be quite unbalanced. I suggest splitting the section into, for example, Case study and Results and discussion.
  (b) The whole section reads more like a thesis chapter than a section in a scientific paper. The text, results, figures, and tables should be better and more clearly presented.

Particular and minor comments:

• The last paragraph of the Introduction should only be used to describe the paper structure. Please fix this part's first sentence: "The rest of this paper is organized as follows: section 1 introduces the related studies on MFD construction". It is non-sensible.
• The work with abbreviations should be improved (e.g., in Abstract and Introduction):
  (a) for example, CSD and ALPR are not defined (out of abstract); what is DS?;
  (b) ALPR is not used in the abstract (besides its definition);
  (c) the paper is almost non-readable due to many used abbreviations – maybe a List of Abbreviations would be helpful.
• There are many typos in the paper (e.g., missing spaces, see lines 147 and 148). Please double-check these issues in the whole manuscript.
• Figure 1 involves a (non-sensible?) mixture of terms with used and non-used abbreviations.
• Fix numbering of Equation referencing (see, e.g., line 149). The equations should be referenced es Eq. (1), not Eq. 1.
• An example of non-understandable (language) issues is in lines 191-193. The paper is full of such sentences and paragraphs.
• Section 4 should preferably be entitled Conclusions (alternatively Conclusions and discussion).

I am not sure if the Conclusions section can involve abbreviations. I rather think that Conclusions (as well as Abstract) should be able to be readable without reading the paper. Please double-check this as it involves quite many abbreviations

Author Response

The newly-added or revised content in the article is in light blue.

Response to Reviewer 1 Comments

 

Point 1: The (English) language of the whole paper should be improved. Extensive proofreading by a native speaker is necessary.

 

Response 1: Thanks for the comments. To improve our writing, firstly, we have conduct a thorough review for English grammar, spelling and wording of our paper. Then, the paper was proofread by a native English editor.

 

Point 2:  Introduction and abstract:

From reading the abstract (and the Introduction), I don't understand the paper's aim, novelty, and contribution. This should be better stated. Therefore, the abstract should be rewritten, and the Introduction should also be rewritten such that it

(a) mentions the importance of the topic,

(b) involves a short state-of-the-art,

(c) explains how the paper contributes to the problem,

(d) describes the structure of the paper.

Such a 4-paragraph approach seems to be relevant for the paper. However, this should be followed by a literature review section.

 

Response 2: Thanks for the comments. The abstract and introduction were rewritten as your suggestion. The specific modifications are shown as follows.

 

“Abstract: Road network traffic management and control of the urban road networks is the main issue to alleviate traffic congestion. The study aims to characterize the traffic flow state of urban road network using Macroscopic Fundamental Diagram (MFD). MFD is a road network-level traffic flow model, the core property of MFD is that the network flow is maximized when network traffic stays at an optimal accumulation state. The property can be used to optimize the temporal and spatial distribution of traffic flow with applications such as gating control. MFD construction is the basis for these MFD-based applications. Although many studies had been conducted to construct MFD, few studies were dedicated to improve the accuracy considering the different re-liability of different sources of data with field data. To this end, this article proposes a MFD construction method using multi-source data based on Dempster-Shafer evidence (DS-evidence) theory considering different data’s reliability. Vehicle trajectory data (VTD) represented by probe vehicle data and cross-section data of traffic flow (CSD) represented by microwave vehicle detectors are selected in this paper. For VTD, a MFD construction method is proposed that considers different penetration of probe vehicles with different origins and destinations (OD) supplemented by automatic license plate recognition (ALPR) data. Both two parameters of MFD are estimated using VTD and CSD, respectively. A fusion method for MFD construction is then proposed through quantifying the reliability of different sources of data for each MFD parameter based on DS-evidence theory. The case study is conducted under real data and simulated data. The results showed that the accuracy of constructed MFD was greatly improved considering different data’s reliability. Specifically, compared to using VTD and CSD respectively, the maximum MFD estimation error is reduced by 22.3% using fusion method. The proposed method provides potentials to support the evaluation of traffic operation and the optimization of signal control scheme for urban traffic network.

Keywords: urban road network; multi-source data; MFD construction; DS-evidence theory

 

1. Introduction

Road network traffic management and control based on traffic supply-demand relationship is an important means to actively alleviate urban road traffic congestion and im-prove the operational efficiency of road traffic systems. Characterize the traffic flow state of urban road network is a fundamental problem to understand the traffic supply-demand relationship. This study aims to characterize the traffic flow state of urban road network using Macroscopic Fundamental Diagram (MFD), because MFD has the following advantages: can dynamically describe the traffic state of road network, can be readily obtained with traffic flow data from existing detectors, etc. The existence of MFD has been confirmed in many cities around the world [1]. The core property of MFD is that the net-work flow is maximized when network traffic stays at an optimal accumulation state. The property can be applied in gating control, traffic safety, congestion toll, etc [2-6].”

…….

However, the location of the cross-sectional vehicle detector has a great impact on the estimation of ATD. Buisson and Ladier [8] divided the CSD into three categories according to the distance between the detector and the downstream signalized intersections, and then constructed MFD using the three types of data. The study found that the closer the detector is to the downstream signalized intersections, the slope of the rising section of the MFD obtained by these detectors is higher. Courbon and Leclercq [9] conducted a comparative analysis of MFD construction based on VTD and CSD, and found that MFD construction depend on the distribution of loop detectors in the road network. Leclercq [10] pointed out that only using CSD to construct MFD is inaccuracy. The reason is that the CSD cannot accurately capture the average traffic speed or density. In addition, Leclercq believes that the accuracy of MFD construction is significantly improved when estimating traffic flow using CSD and estimating traffic speed using VTD compared with the single use of the loop detector data. Tilg et al. [11] systematically evaluated two main analytical approximation methods to derive the MFD for arterial roads and urban networks, and the results revealed that the availability of signal data can improve the analytical approximation of the MFD.

………

In summary, the MFD construction based on CSD is greatly affected by the location of detectors, and that based on VTD has high requirements on the penetration rate and balanced spatial distribution of probe vehicles. The method that estimates different parameters of MFD with different data may lead to the inconsistency between the two parameters. In view of this, a MFD fusion method which can take advantage of different source of data is of urgent need.

Considering the different reliability of estimating ATF and ATD using CSD and VTD, this study developed a MFD construction method based on Dempster-Shafer evidence (DS-evidence) theory using multi-source data. Firstly, the MFD based on CSD and VTD were constructed separately. Secondly, the weight of CSD and VTD for ATF and ATD were separately estimated using DS-evidence theory, and then the fusion MFD was derived and evaluated. The proposed method aims to improve the accuracy of MFD construction by making full use of different sources of data. To this end, the contributions of this paper can be described as follows: 1) a MFD construction method was proposed considering the difference in penetration of probe vehicles with different origins and destinations (OD); 2) a fusion method for MFD construction was proposed considering the different reliability of different sources of data. The paper is organized as follows: Section 1 introduces the related studies on MFD construction; section 2 proposes the methodology for MFD construction; section 3 is the implementation of the proposed method, and the conclusions are in section 4.”

 

Point 3:  Literature review:

(a) The paper should separately involve either a section 2 (or subsection, e.g., 1.1) dealing with the literature review.

(b) Maybe a table involving a survey of main characteristics among literature references would be helpful. Such a table should prove and strengthen the contribution of this paper.

(c) It is quite surprising (or rather worrying) that the literature review does not involve many recent papers (only one paper from 2020+). I am, for example, aware of several (renowned) authors and their recent papers (e.g., prof. Carlos Daganzo or Yicheng Zhang).

(d) The authors claim that the "Macroscopic fundamental diagram provides a new way to optimize…". A new? Just by a brief look, I found the first papers involving "MFD" in the title from 2008 (by Daganzo et al., as cited). Especially this keyword (MFD) and its research progress since 2008 should be discussed in detail!

 

Response 3: Thanks for the comments. First, as the Introduction and Literature review are closed linked, we would like to introduce them together. Then, the Literature review was rewritten, and some recent papers were added in our paper. The modified Literature review can be seen in Response 2.

 

Point 4: Materials and methods:

(a) It is necessary to involve a table (or tables) of, e.g., sets/indices, parameters, and variables used in Section 2. This could either be in Section 2 or in Appendix.

(b) The equations are not clearly presented; for example, during reviewing eq. (5) and (6) I have realized that a scheme of an illustrative example would be helpful (to explain the various notations, parameters, variables, or principles). For example, I don't understand "..

 

Response 4: Thanks for the comments. To improve the readability of our paper, the descriptions of variables used in the paper were listed in Appendix A. Besides, some examples were added to better introduce the equations. An example can be found in the description of eq. (5) and (6).

 

“With the estimated penetration rate, the total travel time and total travel distance are derived by sum of the expanded travel time and travel distance of each OD. According to Edie’s definition [13], the travel time and travel distance are converted into the ATD and ATF, respectively. For example,  is proportional to the sum of expanded travel distance of each OD. The specific method is shown in equation (5) & (6).”

 

Point 5:  Results:

(a) Section 3 is around 12 pages. In the 22 pages long paper, this seems to be quite unbalanced. I suggest splitting the section into, for example, Case study and Results and discussion.

(b) The whole section reads more like a thesis chapter than a section in a scientific paper. The text, results, figures, and tables should be better and more clearly presented.

 

Response 5: Thanks for the comments. To clearly present our paper, the contents in Section 3 were reorganized as your suggestion.

 

Point 6: Particular and minor comments:

The last paragraph of the Introduction should only be used to describe the paper structure. Please fix this part's first sentence: "The rest of this paper is organized as follows: section 1 introduces the related studies on MFD construction". It is non-sensible.

The work with abbreviations should be improved (e.g., in Abstract and Introduction):

(a) for example, CSD and ALPR are not defined (out of abstract); what is DS?;

(b) ALPR is not used in the abstract (besides its definition);

(c) the paper is almost non-readable due to many used abbreviations – maybe a List of Abbreviations would be helpful.

There are many typos in the paper (e.g., missing spaces, see lines 147 and 148). Please double-check these issues in the whole manuscript.

Figure 1 involves a (non-sensible?) mixture of terms with used and non-used abbreviations.

Fix numbering of Equation referencing (see, e.g., line 149). The equations should be referenced es Eq. (1), not Eq. 1.

An example of non-understandable (language) issues is in lines 191-193. The paper is full of such sentences and paragraphs.

Section 4 should preferably be entitled Conclusions (alternatively Conclusions and discussion).

I am not sure if the Conclusions section can involve abbreviations. I rather think that Conclusions (as well as Abstract) should be able to be readable without reading the paper. Please double-check this as it involves quite many abbreviations

 

Response 6: Thanks for the comments. The last paragraph of the Introduction section was modified as your suggestion. To improve the readability of our paper, firstly, some long sentences were modified to short sentences; Secondly, the abbreviations used in the paper were listed in Appendix A. Besides, the proposed abbreviations in Conclusions section were replaced with detailed explanations.

 

 

 

 

 

 

Author Response File: Author Response.pdf

Reviewer 2 Report

Dear Authors,

 

I enjoyed reading your paper.

Still, there are some issues you should deal with.

For instance:

• English language and style issues - Grammarly (https://app.grammarly.com) on default settings detected only for the text block resulting from the concatenation of Title+Abstract+Keywords 12 critical alerts (correctness issues) and 8 more advanced ones, namely: Passive voice misuse (3), Word choice (2), Unclear sentences (1), Misplaced words and phrases (1), and Punctuation in compound/complex sentences (1). This meant an overall score of 63 out of 100 for this sample above. Moreover, since you do not appear to be native English speakers, I suggest a total revision of the English language and style for the entire article using Grammarly or another specialized tool;
• This paper follows the specific structure of the journal, namely:
  Author Information, Abstract, Keywords, Introduction, Materials & Methods, Results, Discussion, Conclusions, etc., as indicated in the corresponding section of the journal: https://www.mdpi.com/journal/sustainability/instructions. However, the first titles for the subsections in section 3 should better reflect the affiliation to the latter (Results).
• You must avoid ending some sections/subsections with figures  (e.g. Figure 7 just before subsection 3.3, Figure 11 just before subsection 3.4, or Figure 13 just before the Discussion section);

• You must ensure that all figures have the required resolution (minimum 1000 pixels width/height, or a resolution of 300 dpi or higher according to the Sustainability Journal’s instructions at: https://www.mdpi.com/journal/sustainability/instructions); Figures 2, 9 and 12 do not appear to meet these resolution requirements;
• The legend of Figure 13 comes out of the page where its content is presented. Try to make it fit on the appropriate page.
• The references to multiple equations/formulas must be provided in a more concise way (e.g. “.equations 1 and 2” instead of “Equation 1-Equation 2” - line 149); the same for 5 and 6, 7 and 8, or 9 and 10;
• I think more contributions in journal papers must be cited in this research both in the Introduction and Discussion sections (I believe that only 22 references, some of which being dissertations and papers at conferences and symposia, are not enough);
• Digital object identifier (DOI) codes for all references are encouraged by the journal. Therefore, you should provide them;

• This paper has so many figures (13) and tables (7). Some of them (not essential for understanding the main content) should be moved to the Appendix section. If not existing, this section must be created.

 

Thank you for your contribution!

 

Sincerely,

D.H.

Author Response

The newly-added or revised content in the article is in light blue.

Response to Reviewer 2 Comments

 

Point 1: English language and style issues - Grammarly (https://app.grammarly.com) on default settings detected only for the text block resulting from the concatenation of Title+Abstract+Keywords 12 critical alerts (correctness issues) and 8 more advanced ones, namely: Passive voice misuse (3), Word choice (2), Unclear sentences (1), Misplaced words and phrases (1), and Punctuation in compound/complex sentences (1). This meant an overall score of 63 out of 100 for this sample above. Moreover, since you do not appear to be native English speakers, I suggest a total revision of the English language and style for the entire article using Grammarly or another specialized tool;

 

Response 1: Thanks for the comments. To improve our writing, firstly, we have conduct a thorough review for English grammar, spelling and wording of our paper. Then, the paper was proofread by a native English editor.

 

Point 2: This paper follows the specific structure of the journal, namely: Author Information, Abstract, Keywords, Introduction, Materials & Methods, Results, Discussion, Conclusions, etc.,

as indicated in the corresponding section of the journal: 

https://www.mdpi.com/journal/sustainability/instructions. However, the first titles for the subsections in section 3 should better reflect the affiliation to the latter (Results).

 

Response 2: Thanks for the comments. To better describe the section 3, the content was reorganized and the titles were also modified.

 

Point 3:  You must avoid ending some sections/subsections with figures  (e.g. Figure 7 just before subsection 3.3, Figure 11 just before subsection 3.4, or Figure 13 just before the Discussion section);

 

Response 3: Thanks for the comments. The sections/subsections ended with figures were modified as your suggestion.

 

Point 4:  You must ensure that all figures have the required resolution (minimum 1000 pixels width/height, or a resolution of 300 dpi or higher according to the Sustainability Journal’s instructions at: https://www.mdpi.com/journal/sustainability/instructions); Figures 2, 9 and 12 do not appear to meet these resolution requirements.

 

Response 4: Thanks for the advice. Figures 2, 9 and 12 were replaced with the ones that meet the resolution requirements.

 

Point 5: The legend of Figure 13 comes out of the page where its content is presented. Try to make it fit on the appropriate page.

 

Response 5: Thanks for your suggestion. The legend of Figure 13 was modified as follows.

(a)

(b)

(a)

Figure 13. The parameter and the scatter plot of constructed MFD using simulated data.(a) The estimated ATD;(b) The estimated ATF;(c) The scatter plot of MFD.

Point 6: The references to multiple equations/formulas must be provided in a more concise way (e.g. “.equations 1 and 2” instead of “Equation 1-Equation 2” - line 149); the same for 5 and 6, 7 and 8, or 9 and 10;

 

Response 6: Thanks for the comments. The references to multiple equations/formulas were modified as your suggestion.

 

Point 7: I think more contributions in journal papers must be cited in this research both in the Introduction and Discussion sections (I believe that only 22 references, some of which being dissertations and papers at conferences and symposia, are not enough);

 

Response 7: Thanks for the comments. The Literature review was rewritten, and some recent papers were added in our paper. The specific modifications are shown as follows.

 

“Road network traffic management and control based on traffic supply-demand relationship is an important means to actively alleviate urban road traffic congestion and im-prove the operational efficiency of road traffic systems. Characterize the traffic flow state of urban road network is a fundamental problem to understand the traffic supply-demand relationship. This study aims to characterize the traffic flow state of urban road network using Macroscopic Fundamental Diagram (MFD), because MFD has the following advantages: can dynamically describe the traffic state of road network, can be readily obtained with traffic flow data from existing detectors, etc. The existence of MFD has been confirmed in many cities around the world [1]. The core property of MFD is that the net-work flow is maximized when network traffic stays at an optimal accumulation state. The property can be applied in gating control, traffic safety, congestion toll, etc [2-6].

…….

However, the location of the cross-sectional vehicle detector has a great impact on the estimation of ATD. Buisson and Ladier [8] divided the CSD into three categories according to the distance between the detector and the downstream signalized intersections, and then constructed MFD using the three types of data. The study found that the closer the detector is to the downstream signalized intersections, the slope of the rising section of the MFD obtained by these detectors is higher. Courbon and Leclercq [9] conducted a comparative analysis of MFD construction based on VTD and CSD, and found that MFD construction depend on the distribution of loop detectors in the road network. Leclercq [10] pointed out that only using CSD to construct MFD is inaccuracy. The reason is that the CSD cannot accurately capture the average traffic speed or density. In addition, Leclercq believes that the accuracy of MFD construction is significantly improved when estimating traffic flow using CSD and estimating traffic speed using VTD compared with the single use of the loop detector data. Tilg et al. [11] systematically evaluated two main analytical approximation methods to derive the MFD for arterial roads and urban networks, and the results revealed that the availability of signal data can improve the analytical approximation of the MFD.

………

In summary, the MFD construction based on CSD is greatly affected by the location of detectors, and that based on VTD has high requirements on the penetration rate and balanced spatial distribution of probe vehicles. The method that estimates different parameters of MFD with different data may lead to the inconsistency between the two parameters. In view of this, a MFD fusion method which can take advantage of different source of data is of urgent need.

Considering the different reliability of estimating ATF and ATD using CSD and VTD, this study developed a MFD construction method based on Dempster-Shafer evidence (DS-evidence) theory using multi-source data. Firstly, the MFD based on CSD and VTD were constructed separately. Secondly, the weight of CSD and VTD for ATF and ATD were separately estimated using DS-evidence theory, and then the fusion MFD was derived and evaluated. The proposed method aims to improve the accuracy of MFD construction by making full use of different sources of data. To this end, the contributions of this paper can be described as follows: 1) a MFD construction method was proposed considering the difference in penetration of probe vehicles with different origins and destinations (OD); 2) a fusion method for MFD construction was proposed considering the different reliability of different sources of data. The paper is organized as follows: Section 1 introduces the related studies on MFD construction; section 2 proposes the methodology for MFD construction; section 3 is the implementation of the proposed method, and the conclusions are in section 4.”

 

Point 8: Digital object identifier (DOI) codes for all references are encouraged by the journal. Therefore, you should provide them;

 

Response 8: Thanks for the comments. The Digital object identifier (DOI) codes for most of the references were added.

 

Point 9: This paper has so many figures (13) and tables (7). Some of them (not essential for understanding the main content) should be moved to the Appendix section. If not existing, this section must be created.

 

Response 9: Thanks for the comments. Figure 13 and tables 7 are important evaluation results for the proposed method,therefore, therefore, they were not moved to the Appendix section.

 

Author Response File: Author Response.pdf

Reviewer 3 Report

The manuscript presents experimental results of utilising a numerical method for traffic modelling. The analysis seems to be essential and useful for potential practical implementation. Nevertheless, while reading the material, one can identify some minor shortcomings needing improvement before publishing the article.

Having read the article, one can ask why the study is not referred to the algorithms and methodology used by the leading geo-services providers? Traffic simulations and the relevant information about possible traffic jams is the subject of numerous map portals in different countries. 

Both the abstract and the introduction contain some repetitions and linguistic inaccuracies that need corrections. I recommend reading out the text, trying to find some synonyms or rephrasing sentences so that they sound clearer. Also, please watch missing spaces (ex. lines 66, 148, 152, 153, 154 and so on), commas etc.

Line 149: Please avoid repetitions - instead of 'Equation 1 - Equation 2' please simply use 'Equations 1 & 2'. The remark also refers to the forthcoming passages (as the lines 184, 194, 210, 213 etc.). 

Line 169: A general remark to the equations: in case they represent the authors' findings, they should be additionally explained. If, however, not - they all need accompanying references. 

Is equation 15 the authors' invention? Had the model used been somehow evaluated before the test was performed?

Figure 2 needs a scalebar. In case the geographical location is important, the material needs to be completed by adding a general overview map showing the test site. 

 

Figure 6: both charts look practically the same. In line 457, however, there is a statement about the functional similarity of all such situations. Please clarify it or complete figure 6 with a relevant comment.

Figure 12 needs a scalebar. Also, the information about the GIS software used is mandatory.

Is the invented approach universal, or is it just limited to Kunshan City, Jiangsu Province? Both possible answers (yes or no plus relevant explanations) need some comments. 

The discussion lacks a compelling tagline summarising the taken research. Are the obtained results successful or not? Does the study need some further development? What is the next step planned? 

Should the authors provide the necessary corrections, the text needs resubmission for the next reviewing cycle. Good luck!

 

 

Author Response

The newly-added or revised content in the article is in light blue.

Response to Reviewer 3 Comments

 

Point 1:  Having read the article, one can ask why the study is not referred to the algorithms and methodology used by the leading geo-services providers? Traffic simulations and the relevant information about possible traffic jams is the subject of numerous map portals in different countries. 

 

Response 1: Thanks for the comments. The references about algorithms and methodology used by geo-services providers can be found in the part of the data preprocessing for vehicle trajectory data. As it is not the key problem in our study, it was briefly introduced in Section 3.1.1.

 

Point 2:  Both the abstract and the introduction contain some repetitions and linguistic inaccuracies that need corrections. I recommend reading out the text, trying to find some synonyms or rephrasing sentences so that they sound clearer. Also, please watch missing spaces (ex. lines 66, 148, 152, 153, 154 and so on), commas etc.

 

Response 2: Thanks for the comments. To improve our writing, we have conduct a thorough review for English grammar, spelling and wording of our paper.

 

Point 3:  Line 149: Please avoid repetitions - instead of 'Equation 1 - Equation 2' please simply use 'Equations 1 & 2'. The remark also refers to the forthcoming passages (as the lines 184, 194, 210, 213 etc.). 

 

Response 3: Thanks for the comments. The references to multiple equations/formulas were modified as your suggestion.

 

Point 4:  Line 169: A general remark to the equations: in case they represent the authors' findings, they should be additionally explained. If, however, not - they all need accompanying references. 

 

Response 4: Thanks for the comments. The equations which are not our findings were added with references.

 

Point 5:  Is equation 15 the authors' invention? Had the model used been somehow evaluated before the test was performed?

 

Response 5: Thanks for the comments. Equation 15 is not our invention, the reference [14] was added in the description of equation 15. In the reference, the model was used to estimate travel time. And it was evaluated through a comparison with the alternative methods, the results revealed that the model is shown to be able to consistently generate improved performance under varying traffic conditions.

 

Point 6:  Figure 2 needs a scale bar. In case the geographical location is important, the material needs to be completed by adding a general overview map showing the test site. 

 

Response 6: Thanks for the comments. The scale bar was added to figure 2 as your suggestion.

 

Point 7:  Figure 6: both charts look practically the same. In line 457, however, there is a statement about the functional similarity of all such situations. Please clarify it or complete figure 6 with a relevant comment.

 

Response 7: Thanks for the comments. Figure 6 depicts the average penetration rate of probe vehicles in each period for weekdays and weekends, respectively. Line 457 is the description of the constructed MFD for weekdays and weekends, respectively. The difference of penetration rate is one of the key factors that lead to the significant difference in MFD between weekdays and weekends (The specific description can be found in Section 3.2). To clarify this, some sentences were added for the description of figure 6.

 

Point 8:  Figure 12 needs a scalebar. Also, the information about the GIS software used is mandatory.

 

Response 8: Thanks for the comments. Figure 12 is the simulated road network of the test site using a traffic simulation software Q-PARAMICS (Quadstone Parallel Microscopic Simulator). The road network was draw of a scale of 1:1 with actual road network in figure 2. And the scale bar is almost the same as which in figure 2. The specific introduction can be found in section 3.3.

 

Point 9:  Is the invented approach universal, or is it just limited to Kunshan City, Jiangsu Province? Both possible answers (yes or no plus relevant explanations) need some comments. 

 

Response 9: Thanks for the comments. The proposed approach can be applied in any city with mentioned data. And the accuracy of constructed MFD is subjected to the probe vehicles’ penetration rate, the fixed detectors’ coverage rate and the spatial equilibrium degree of probe vehicles and fixed detectors, etc.

 

Point 10:  The discussion lacks a compelling tagline summarizing the taken research. Are the obtained results successful or not? Does the study need some further development? What is the next step planned? 

 

Response 10: Thanks for the comments. The discussion section was rewritten as your suggestion. The specific modifications are shown as follows.

 

“Taking the congested area of Kunshan City, Jiangsu Province as the test site, the results for actual and simulated scenarios were analyzed, and the performance of the proposed method was evaluated. The results of case study show that the changing trends of MFD in different periods under the three data are the same, and the values are similar, which confirms the reliability of the proposed method. The results also show the high accuracy (with a minimum error of 7.7%) of constructed MFD under multi-source data. For the same scenario, the MFD construction error of the fusion method has a maximum reduction of 11% compared with which only using cross-sectional traffic flow data, and has a maximum reduction of 22.3% compared with which only using vehicle trajectory data, and has a maximum reduction of 8.1% compared with those based on the MFD construction through estimates different parameters using different data. In summary, the proposed method can greatly improve the accuracy of constructed MFD through combining the advantages of two types of data. It provides potentials to support the evaluation of traffic operation and the optimization of signal control scheme for urban traffic network. However, due to the limitations of data’s availability, the accuracy of proposed method can’t be evaluated in cities with severe congestion (such as Beijing) whose descending part of MFD would present. In future, the impact factors that affect the accuracy of constructed MFD will be investigated through comparative analysis. Also, more field validation works are needed to valid the effectiveness of the proposed method.”

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

This is a re-review of the article.

As reported in the previous round of review, I have found some drawbacks. The major weaknesses were reported. The authors made some corrections to the manuscript, and they even pretended, in their response to Reviewers, that they incorporated all my suggestions (or at least most of them). Unfortunately, I have serious doubts about this as the most of them were not properly addressed in the revision:

• The abstract was rewritten. However, it does not follow guidelines for authors of the Sustainability journal (the maximum number of words should be about 200).
• The introduction was rewritten. However, the suggestion to reduce the number of paragraphs was ignored.
• The suggestion about extending the literature review on a review table was also ignored (similarly to the suggestion for the separate (sub-)section).
• Even if the authors created Appendix A, they have mixed the list of abbreviations and the notation of variables, parameters, etc., into one file. They even omitted most of the abbreviations.
• Other major or minor comments and suggestions, e.g., about
  (a) the length of Section 3,
  (b) the non-sensible paragraph on the paper structure,
  (c) the Figure 1,
  (c) an illustrative scheme by eq. (5) and (6))

were also ignored.

To sum up, I would recommend major revision for a local journal with a lower impact factor. I am, however, not comfortable recommending the acceptance of the manuscript to Sustainability. As such, I suggest rejection.

Author Response

Bing Wang,

Professor

School of Traffic &Transportation Engineering,

Xinjiang University, Urumqi, Xinjiang, China

[email protected]

 

 

May, 18, 2022

Sustainability

 

 

Dear Reviewer 1:

 

I am writing to you regarding the modifications made to our paper (1707041, “A MFD Construction Method Considering Multi-Source Data Reliability for Urban Road Networks”) in light of your comments.

 

All of the comments from you were carefully studied and followed when revising the paper. The details about our responses to your comments are summarized in the attachment at the end of this letter for your convenience.

 

We truly appreciate your constructive comments and believe that the revised paper is now clearer and more useful to researchers and practitioners. If you have further questions, please feel free to contact me at [email protected] .

 

Sincerely Yours,

 

 

Bing Wang

--------------------------------

Bing Wang,

Professor

#777 Hua rui Street

School of Traffic &Transportation Engineering,

Xinjiang University, Urumqi, 830017, China

Email: [email protected]

 

 

Attachment: Authors’ Response to Reviewers

The Authors’ Response are in red. The newly-added or revised content in the article is in light blue.

 

Response to Reviewer 1 Comments

 

Point 1: The abstract was rewritten. However, it does not follow guidelines for authors of the Sustainability journal (the maximum number of words should be about 200).

 

Response 1: Thanks for the comments. The number of words of the abstract was modified to 225, the modified abstract is shown as follows.

 

“Abstract: Road network traffic management and control is the main issue to alleviate urban traffic congestion. The study aims to characterize the traffic flow state of urban road networks using the Macroscopic Fundamental Diagram (MFD) to support area traffic control. The core property of MFD is that the network flow is maximized when network traffic stays at an optimal accumulation state. The property can be used to optimize the temporal and spatial distribution of traffic flow with applications such as gating control. MFD construction is the basis for these MFD-based applications. Although many studies had been conducted to construct MFD, few studies were dedicated to improving the accuracy considering the reliability of different sources of data. To this end, this article proposes a MFD construction method using multi-source data based on Dempster-Shafer evidence (DS-evidence) theory considering different data’s reliability. First, the MFD was con-structed using VTD and CSD, separately. Then, the fused MFD was derived by quantifying the reliability of different sources of data for each MFD parameter based on DS-evidence theory. The results under real data and simulated data showed that the accuracy of constructed MFD was greatly improved considering different data’s reliability (the maximum MFD estimation error is reduced by 22.3%). The proposed method provides potentials to support the evaluation of traffic operation and the optimization of signal control schemes for urban traffic networks.”

 

Point 2: The introduction was rewritten. However, the suggestion to reduce the number of paragraphs was ignored.

 

Response 2: Thanks for your valuable suggestion. The introduction was rewritten and the number of paragraphs was also reduced as your suggestion.

 

“Road network traffic management and control based on traffic supply-demand relationship is an important means to actively alleviate urban traffic congestion and improve the operational efficiency of road traffic systems. Characterizing the traffic flow state of urban road networks is a fundamental problem to understand the traffic supply-demand relationship. This study aims to characterize the traffic flow state of urban road networks using the Macroscopic Fundamental Diagram (MFD) because MFD has the following advantages: it can dynamically describe the traffic state of the road network, and it can be readily obtained with traffic flow data from existing detectors, etc. The existence of MFD has been confirmed in many cities around the world [1]. The core property of MFD is that the network traffic flow is maximized when network traffic stays at an optimal accumulation state (the network traffic flow remains maximum with increasing network traffic density). The property can be applied in gating control, traffic safety, congestion toll, etc. [2-6]. The construction of MFD is the basis of MFD-based applications.

Generally, there are three types of data to construct MFD. Cross-section data of traffic flow (CSD), vehicle trajectory data (VTD), and both two sources of data. The general method to construct MFD is weighted average method for CSD and Edie's method for VTD. However, the MFD construction based on CSD is greatly affected by the location of detectors, and that based on VTD has high requirements on the penetration rate and balanced spatial distribution of probe vehicles. Few of the Multi-source of data-based method considering the different reliability of different sources of data. And the method that estimates different parameters of MFD with different data may lead to the inconsistency between the two parameters. In view of this, a MFD fusion method that can take advantage of different sources of data is of urgent need.

Considering the different reliability of estimating ATF and ATD using CSD and VTD, this study developed a MFD construction method based on the Dempster-Shafer evidence (DS-evidence) theory using multi-source data. Firstly, the MFD based on CSD and VTD were constructed separately. Secondly, the weight of CSD and VTD for ATF and ATD were separately estimated using DS-evidence theory, and then the fusion MFD was derived and evaluated. The proposed method aims to improve the accuracy of MFD construction by making full use of different sources of data. To this end, the contributions of this paper can be described as follows: 1) a MFD construction method was proposed considering the difference in penetration of probe vehicles with different origins and destinations (OD); 2) a fusion method for MFD construction was proposed considering the different reliability of different sources of data.

The rest of our paper is organized as follows: section 2 is the literature review of MFD construction; section 3 proposes the methodology for MFD construction; section 4 is the case study; section 5 is the implementation of the proposed method, and the conclusions and discussion are in section 6. “

 

Point 3:  The suggestion about extending the literature review on a review table was also ignored (similarly to the suggestion for the separate (sub-)section).

 

Response 3: Thanks for your valuable suggestion. The review table of literature review was added in Section 2 (Table 1), and the literature review was separated as section 2.

Table 1. Representative Research Summary

The type of data	Data sources	Method	Representative research
Cross-section data of traffic flow	Loops and microwave detectors	Weighted average method	Geroliminis and Daganzo [1]
	VISSIM simulation	Weighted average method	Ortigosa and Menendez [7]
	Loop detectors	Unweighted average method	Buisson and Ladier [8]
	Simulation	Weighted average method	Courbon and Leclercq [9]
	Simulation	Weighted average method	Leclercq [10]
Vehicle trajectory data	Taxis’ GPS	Edie's method	Geroliminis and Daganzo [1]
	Simulation	Edie's method	Courbon and Leclercq [9]
	Simulation	Edie's method	Leclercq [10]
	Mobile Probe Data	Edie's method	Nagle and Gayah [12]
Multi-source of data	Loop detectors data and floating car data	Weighted average method, Edie's method and fusion algorithm considering network coverage of each data type	Ambühl and Menendez [13]
	Loop detector data and Floating car data from simulation software	Weighted average method, Edie's method and Bayesian fusion method	Saffari et al. [16]
	Loop detectors and Taxis’ GPS	Weighted average method, Edie's method	Ji et al. [20]

 

Point 4:  Even if the authors created Appendix A, they have mixed the list of abbreviations and the notation of variables, parameters, etc., into one file. They even omitted most of the abbreviations.

 

Response 4: Thanks for the comments. The abbreviations and the notation of variables were listed separately, and all related abbreviations were included.

 

Point 5:  Section 3 is around 12 pages. In the 22 pages long paper, this seems to be quite unbalanced. I suggest splitting the section into, for example, Case study and Results and discussion.

 

Response 5: Thanks for the comments. Section 3 was split into Case study (Section 4) and Results (Section 5).

 

Point 6: The last paragraph of the Introduction should only be used to describe the paper structure. Please fix this part's first sentence: "The rest of this paper is organized as follows: section 1 introduces the related studies on MFD construction". It is non-sensible.

 

Response 6: Thanks for the comments. The paper structure was reorganized as follows.

 

“The rest of our paper is organized as follows: section 2 is the literature review of MFD construction; section 3 proposes the methodology for MFD construction; section 4 is the case study; section 5 is the implementation of the proposed method, and the conclusions and discussion are in section 6.”

 

Point 7: Figure 1 involves a (non-sensible?) mixture of terms with used and non-used abbreviations.)

 

Response 7: Thanks for the comments. To better explain the framework of our study, most of the abbreviations were replaced with detail descriptions. The Figure 1 was modified as follows.

 

Point 8: The equations are not clearly presented; for example, during reviewing eq. (5) and (6) I have realized that a scheme of an illustrative example would be helpful (to explain the various notations, parameters, variables, or principles). For example, I don't understand ".

 

Response 8: Thanks for the comments. To better introduce equation (5) and (6), the description of eq. (5) and (6) was modified as follows.

 

“With the estimated penetration rate, the total travel time and total travel distance were derived by the sum of the expanded travel time and travel distance of each OD. According to Edie’s definition [21], the travel time and travel distance were converted into the ATD and ATF, respectively. For example, the network traffic flow during period t is equal to the number of times all vehicle running on the road network during period t.”

Author Response File: Author Response.pdf

Reviewer 2 Report

Dear Authors,

You solved most of the issues pointed out in the first round of revisions.
I think the paper is now ready for being published.

Sincerely,
D.H.

Author Response

Bing Wang,

Professor

School of Traffic &Transportation Engineering,

Xinjiang University, Urumqi, Xinjiang, China

[email protected]

 

 

May, 18, 2022

Sustainability

 

 

Dear Reviewer 2:

 

I am writing to you regarding the modifications made to our paper (1707041, “A MFD Construction Method Considering Multi-Source Data Reliability for Urban Road Networks”) in light of your comments.

 

All of the comments from you were carefully studied and followed when revising the paper.The details about our responses to your comments are summarized in the attachment at the end of this letter for your convenience.

 

We truly appreciate your constructive comments and believe that the revised paper is now clearer and more useful to researchers and practitioners. If you have further questions, please feel free to contact me at [email protected] .

 

Sincerely Yours,

 

 

Bing Wang

--------------------------------

Bing Wang,

Professor

#777 Hua rui Street

School of Traffic &Transportation Engineering,

Xinjiang University, Urumqi, 830017, China

Email: [email protected]

 

Attachment: Authors’ Response to Reviewers

The Authors’ Response are in red. The newly-added or revised content in the article is in light blue.

 

Response to Reviewer 2 Comments

 

Thanks for your acknowledgment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Dear Authors,

Thank you very much for your responses to my comments and suggestions. I appreciate them, and I think your work should be published. However, I have still found some minor inaccuracies that should be levelled before further text processing.

Line 52 contains some non-English letters - please either translate the text into English or remove it.

Figure 2 - the newly added scale bar is inappropriate. Only round decimal values are allowed (the distances should be expressed in meters - international SI units are obligatory!). Using fractions (like 0.0075 etc.) for scale bars is unacceptable in cartography! Please refer to any cartographic workbook for more details about the visualisation methods.

If you use cartographic coordinates obtained from GPS, you'd better scale your figures by adding a cartographic grid. Hence, we don't know where your test field is located!

Figure 12 consists of two screenshots presenting the road network used for tests. Either please explain the reason for showing both pictures or remove one of them. If you aimed to show different scales, please explain them in the figure description.

The cartographic part of the article is the weakest one and must be improved before the text is internationally published. Good luck!

 

 

Author Response

Bing Wang,

Professor

School of Traffic &Transportation Engineering,

Xinjiang University, Urumqi, Xinjiang, China

[email protected]

 

 

May, 18, 2022

Sustainability

 

 

Dear Reviewer 3:

 

I am writing to you regarding the modifications made to our paper (1707041, “A MFD Construction Method Considering Multi-Source Data Reliability for Urban Road Networks”) in light of your comments.

 

All of the comments from you were carefully studied and followed when revising the paper. The details about our responses to your comments are summarized in the attachment at the end of this letter for your convenience.

 

We truly appreciate your constructive comments and believe that the revised paper is now clearer and more useful to researchers and practitioners. If you have further questions, please feel free to contact me at [email protected] .

 

Sincerely Yours,

 

 

Bing Wang

--------------------------------

Bing Wang,

Professor

#777 Hua rui Street

School of Traffic &Transportation Engineering,

Xinjiang University, Urumqi, 830017, China

Email: [email protected]

 

 

Attachment: Authors’ Response to Reviewers

The Authors’ Response are in red. The newly-added or revised content in the article is in light blue.

 

Response to Reviewer 3 Comments

 

Point 1: Line 52 contains some non-English letters - please either translate the text into English or remove it.

 

Response 1: Thanks for the comments. The non-English letters in Line 52 were deleted (by the way, this error belongs to an intermediate state during revising).

 

Point 2: Figure 2 - the newly added scale bar is inappropriate. Only round decimal values are allowed (the distances should be expressed in meters - international SI units are obligatory!). Using fractions (like 0.0075 etc.) for scale bars is unacceptable in cartography! Please refer to any cartographic workbook for more details about the visualisation methods. If you use cartographic coordinates obtained from GPS, you'd better scale your figures by adding a cartographic grid. Hence, we don't know where your test field is located!

 

Response 2: Thanks for the comments. The newly added scale bar was modified as follows.

 

 

Point 3:  Figure 12 consists of two screenshots presenting the road network used for tests. Either please explain the reason for showing both pictures or remove one of them. If you aimed to show different scales, please explain them in the figure description.

 

Response 3: Thanks for the comments. One of the two screenshots presenting the road network used for tests of Figure 12 was deleted (by the way, this error belongs to an intermediate state during revising).

 

 

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

This is a 3rd review of the manuscript „A MFD Construction Method Considering Multi-Source Data 2 Reliability for Urban Road Networks“.

The authors have addressed the very most of my comments and suggestions now. In my eyes, the paper can be accepted after a few minor revisions listed below.

• Table 1:
  (a) It is unclear which articles/works belong to which „The typo of data“ column.
  (b) There should be one more row representing „This article“; this should nicely approve the novelty of this manuscript.
• Figure 1: why some abbreviations, e.g., VTD and CSD, are not used?
• Appendix (both A and B): It should include information about its content (e.g., via a table caption or via a subsection title): for example, Appendix A: List of abbreviations, Appendix B: Used notation.

Author Response

Bing Wang,

Professor

School of Traffic &Transportation Engineering,

Xinjiang University, Urumqi, Xinjiang, China

[email protected]

 

 

May, 18, 2022

Sustainability

 

 

Dear Reviewer 1:

 

I am writing to you regarding the modifications made to our paper (1707041, “A MFD Construction Method Considering Multi-Source Data Reliability for Urban Road Networks”) in light of your comments.

 

All of the comments from you were carefully studied and followed when revising the paper. The details about our responses to your comments are summarized in the attachment at the end of this letter for your convenience.

 

We truly appreciate your constructive comments and believe that the revised paper is now clearer and more useful to researchers and practitioners. If you have further questions, please feel free to contact me at [email protected] .

 

Sincerely Yours,

 

 

Bing Wang

--------------------------------

Bing Wang,

Professor

#777 Hua rui Street

School of Traffic &Transportation Engineering,

Xinjiang University, Urumqi, 830017, China

Email: [email protected]

 

Attachment: Authors’ Response to Reviewers

The Authors’ Response are in red. The newly-added or revised content in the article is in light blue.

 

Response to Reviewer 1 Comments

 

Point 1: Table 1: (a) It is unclear which articles/works belong to which „The typo of data“ column. (b) There should be one more row representing „This article“; this should nicely approve the novelty of this manuscript.

 

Response 1: Thanks for your valuable suggestions. To clarify the type of data and emphasize our work, the table was modified as follows.

Table 1. Summary of representative research

The type of data	Data sources	Method	Representative research
Cross-section data of traffic flow	Loops and microwave detectors	Weighted average method	Geroliminis and Daganzo [1]
	VISSIM simulation	Weighted average method	Ortigosa and Menendez [7]
	Loop detectors	Unweighted average method	Buisson and Ladier [8]
	Simulation	Weighted average method	Courbon and Leclercq [9]
	Simulation	Weighted average method	Leclercq [10]
Vehicle trajectory data	Taxis’ GPS	Edie's method	Geroliminis and Daganzo [1]
	Simulation	Edie's method	Courbon and Leclercq [9]
	Simulation	Edie's method	Leclercq [10]
	Mobile Probe Data	Edie's method	Nagle and Gayah [12]
Multi-source of data	Loop detectors data and floating car data	Weighted average method, Edie's method and fusion algorithm considering network coverage of each data type	Ambühl and Menendez [13]
	Loop detector data and Floating car data from simulation	Weighted average method, Edie's method and Bayesian fusion method	Saffari et al. [16]
	Loop detectors and Taxis’ GPS	Weighted average method, Edie's method	Ji et al. [20]
	Microwave vehicle detectors, Automatic License Plate Recognition Equipment, Taxis’ GPS	Weighted average method, Edie's method and DS-evidence theory fusion method	This article

 

Point 2: Figure 1: why some abbreviations, e.g., VTD and CSD, are not used?

 

Response 2: Thanks for the comments. To increase the readability of the framework in Figure 1, some abbreviations was modified as detailed descriptions. However, as VTD and CSD frequently presented at the following sections, they were replaced as abbreviations.

 

Point 3: Appendix (both A and B): It should include information about its content (e.g., via a table caption or via a subsection title): for example, Appendix A: List of abbreviations, Appendix B: Used notation.

 

Response 3: Thanks for the comments. The content of Appendix was added as follows.

 

“Appendix A: List of abbreviations

Appendix B: List of used notations with their meaning”

 

Author Response File: Author Response.pdf

Reviewer 3 Report

Dear Authors,

Thank you for your extensive and professional responses to my questions and concerns. I appreciate them, and I consider your manuscript much clear now. I have read your newly submitted version, and I noticed all your input changes. Nevertheless, there is still one crucial shortcoming needing improvement before publishing. 

Figure 2 - you added a scale bar; however, it is unfortunately incorrect. Please note that only round, decimal values are allowed - in your case, it must be, e.g.:

0 - 100 - 200 - 500 - 1000 m

The cartographers can not accept the current scalebar!

Hence, I do not have any more objections, so please submit the article again if you have the figure improved.  I wish you good luck!

Author Response

Bing Wang,

Professor

School of Traffic &Transportation Engineering,

Xinjiang University, Urumqi, Xinjiang, China

[email protected]

 

 

May, 18, 2022

Sustainability

 

 

Dear Reviewer 3:

 

I am writing to you regarding the modifications made to our paper (1707041, “A MFD Construction Method Considering Multi-Source Data Reliability for Urban Road Networks”) in light of your comments.

 

All of the comments from you were carefully studied and followed when revising the paper. The details about our responses to your comments are summarized in the attachment at the end of this letter for your convenience.

 

We truly appreciate your constructive comments and believe that the revised paper is now clearer and more useful to researchers and practitioners. If you have further questions, please feel free to contact me at [email protected] .

 

Sincerely Yours,

 

 

Bing Wang

--------------------------------

Bing Wang,

Professor

#777 Hua rui Street

School of Traffic &Transportation Engineering,

Xinjiang University, Urumqi, 830017, China

Email: [email protected]

 

Attachment: Authors’ Response to Reviewers

The Authors’ Response are in red. The newly-added or revised content in the article is in light blue.

 

Response to Reviewer 3 Comments

 

Point 1: Figure 2 - you added a scale bar; however, it is unfortunately incorrect. Please note that only round, decimal values are allowed - in your case, it must be, e.g.:

0 - 100 - 200 - 500 - 1000 m

The cartographers can not accept the current scalebar!

 

Response 1: Thanks for the comments. The modified scale bar was shown as follows.

 

 

Author Response File: Author Response.pdf