Application of an Optimized PSO-BP Neural Network to the Assessment and Prediction of Underground Coal Mine Safety Risk Factors
Round 1
Reviewer 1 Report
This paper aims to develop an efficient model for assessing and predicting safety risk factors in underground coal mines using existing data from the Xiaonan coal mine. An improved hybrid PSO-BP neural network was developed for the evaluation and prediction of safety risk factors in underground coal mines. The empirical analysis showed that the PSO-BP neural network model was the most reliable and effective method for the assessment and prediction of safety risk in underground coal mines.
The paper is well written, and the figures and tables are easy to understand, providing a clear overview of the study. The authors have made an important contribution to the field of mining safety by proposing a reliable and efficient method for assessing and predicting safety risk factors in underground coal mines. Here are some comments and suggestions for further revision before it can be accepted for publication.
As introduced in Section 2.2, PSO algorithm is a population-based metaheuristic optimization algorithm inspired by the social behaviour of birds flocking or fish schooling. It has been widely used to solve a variety of optimization problems. More relevant work should be presented, such as: doi: 10.3390/app12178392; doi: 10.1007/s00170-020-06394-4.
In Section 3.2, the authors list the parameters of PSO algorithm, including particle number, maximum iterations, inertia weight, etc. The authors are suggested to clarify how to determine the values of these parameters, or have they tested the performance of different parameter settings?
Table 4 and Figure 7 demonstrate that PSO-BP neural network has better performance than the traditional BP neural network, which is great. However, it is recommended that in-depth analysis should be conducted and presented to explain why the proposed model is better, or what are the unique advantages related to this method?
Author Response
Response to reviewer 1
Reviewer 1
Comment
This paper aims to develop an efficient model for assessing and predicting safety risk factors in underground coal mines using existing data from the Xiaonan coal mine. An improved hybrid PSO-BP neural network was developed for the evaluation and prediction of safety risk factors in underground coal mines. The empirical analysis showed that the PSO-BP neural network model was the most reliable and effective method for the assessment and prediction of safety risk in underground coal mines.
The paper is well written, and the figures and tables are easy to understand, providing a clear overview of the study. The authors have made an important contribution to the field of mining safety by proposing a reliable and efficient method for assessing and predicting safety risk factors in underground coal mines. Here are some comments and suggestions for further revision before it can be accepted for publication.
Response
Thank you. We have had your comments carefully considered and addressed in the report as required.
Comment
As introduced in Section 2.2, PSO algorithm is a population-based metaheuristic optimization algorithm inspired by the social behaviour of birds flocking or fish schooling. It has been widely used to solve a variety of optimization problems. More relevant work should be presented, such as: doi: 10.3390/app12178392; doi: 10.1007/s00170-020-06394-4.
Response
We have improved upon the presentation of this algorithm by adding relevant materials and following the structure of the referenced materials you provided.
Comment
In Section 3.2, the authors list the parameters of PSO algorithm, including particle number, maximum iterations, inertia weight, etc. The authors are suggested to clarify how to determine the values of these parameters, or have they tested the performance of different parameter settings?
Response
The parameters of the PSO algorithm were selected by performing a sensitivity analysis to evaluate the effect of different parameter values on the optimization result, as mentioned in the discussion under Section 3.1.3, and selecting a reasonable value based on the trade-off between exploration and exploitation.
Comment
Table 4 and Figure 7 demonstrate that PSO-BP neural network has better performance than the traditional BP neural network, which is great. However, it is recommended that in-depth analysis should be conducted and presented to explain why the proposed model is better, or what are the unique advantages related to this method?
Response
As you rightly indicated, we have in accordance included the advantages of this model in the discussion of the results section. This can be found in lines 441 – 445.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Reviewer Comments
Comments to the author:
In this paper, analysis of ‘Application of an optimized PSO-BP neural network to the as- 2 sessment and prediction of underground coal mine safety risk 3 factors’ was debated. The BP neural network is selected as the evaluation method, and a model for evaluating the safety risk in underground coal mining is developed based on the optimized PSO-BP neural network. Following comments should be considered.
1. In title, author used full stop ‘.’ And should be removed.
2. The abbreviations BP, PSO-BP should be addressed with full name in the abstract section.
3. The list of abbreviations should be mentioned at the end of manuscript.
4. The abstract should be not more than 200 words.
5. The line legend of figure one must be corrected.
6. Verify the results of table 1 and table 2.
7. The Introduction should make a compelling case for why the study is useful along with a clear statement of its novelty or originality by providing relevant information and providing answers to basic questions such as:
a. What is already known in the open literature?
b. What is missing (i.e., research gaps)?
c. What needs to be done, why and how?
8. State the motivation and novelty of the work.
9. The conclusion section should be inform of main points rather than a graph.
Comments for author File: 
Comments.pdf
Author Response
Response to reviewer 2
Reviewer 2
In this paper, analysis of ‘Application of an optimized PSO-BP neural network to the as- 2 sessment and prediction of underground coal mine safety risk 3 factors’ was debated. The BP neural network is selected as the evaluation method, and a model for evaluating the safety risk in underground coal mining is developed based on the optimized PSO-BP neural network. Following comments should be considered.
Comment 1
In title, author used full stop ‘.’ And should be removed.
Response
The full stop has been removed accordingly.
Comment 2
The abbreviations BP, PSO-BP should be addressed with full name in the abstract section.
Response
These abbreviations have been addressed in the abstract as instructed.
Comment 3
The list of abbreviations should be mentioned at the end of manuscript.
Response
The list of abbreviations has been inserted at the end of the report as requested.
Comment 4
The abstract should be not more than 200 words.
Response
The abstract has been modified accordingly and the number of words has been reduced to exactly 200.
Comment 5
The line legend of figure one must be corrected.
Response
The line was included just to show the slope of the graph which upon having a second thought, we are of the believe that its role is insignificant. Accordingly, we have taken out the line from the graph.
Comment 6
Verify the results of table 1 and table 2.
Response
The results in table 1 and 2 have been verified accordingly.
Comment 7
The Introduction should make a compelling case for why the study is useful along with a clear statement of its novelty or originality by providing relevant information and providing answers to basic questions such as:
- What is already known in the open literature?
- What is missing (i.e., research gaps)?
- What needs to be done, why and how?
Comment 8
State the motivation and novelty of the work.
Response
We wish to merge comments 7 and 8 and provide the response as follows. We wish to thank you for the pertinent questions and comments. There have been reports of safety risk factors that have in many instances led to accidents. Loss of lives, disabilities and many others have been the end result. For this reason, we took the initiative to gather data on risk factors in the Xiaonan coal mine to aid the promulgation of an efficient model that can be used to assess and predict risk factors in underground coal mines. To the best of our knowledge, this is the first study to have proposed the PSO-BP neural network model for the assessment of safety risk factors in underground coal mines. These have been included in the last paragraph of the introduction.
Comment 9
The conclusion section should be informed of main points rather than a graph.
Response
Accordingly, the conclusion has been modified to include the main ideas in the manuscript.
Author Response File: Author Response.docx
Reviewer 3 Report
Comments and Suggestions for Authors
In this paper, you implemented PSO-BP neural network model to assess and predict underground coal mine safety risk factors. I have a few observations that can help improve the quality of the paper.
1- Statistical information about data is missing in the paper. Please give detailed information about the data in the paper.
2- How did you select PSO tuning parameters in Table 1? Please explain it clearly.
3- There are some minor typos in the manuscript. The manuscript should be carefully re-examined and these errors should be corrected.
4- In Line 220, you should correct the interval as “[0, 1]”.
5- In Line 226, you should correct the interval as “[0.4, 0.9]”. Also, why did you choose this range for the inertia weight? Please define it clearly.
6- What is the objective function used in the study? Please define it as an equation.
7- The first letters of the names of tables and figures in the text should be capitalized like Table 1 etc.
8- Please specify the limitation(s) of the study in the paper.
9- In Table 2, you have presented MSE, MAPE, and R^2 values for 10 models for PSO-BP neural network. What are these 10 models? What are the differences of these 10 models? Please specify clearly.
10- There is no detailed information about GA-BP neural network model used in the study. You should explain this model clearly and give detailed information for this model like the design of the GA-BPNN model, etc.
11- The similarity score should be rechecked and reduced.
Author Response
Response to reviewer 3
Reviewer 3
In this paper, you implemented PSO-BP neural network model to assess and predict underground coal mine safety risk factors. I have a few observations that can help improve the quality of the paper.
Comment 1
Statistical information about data is missing in the paper. Please give detailed information about the data in the paper.
Response
The data used in this report were gathered from the Xiaonan coal mine company. A questionnaire survey mechanism was used to gather a total of 329 datasets. A brief introduction regarding the data used in this report were introduced in the first paragraph in section 4. However, we have attached a data availability statement in the report to ensure clarity and its availability upon request.
Comment 2
How did you select PSO tuning parameters in Table 1? Please explain it clearly.
Response
The parameters of the PSO algorithm were selected by performing a sensitivity analysis to evaluate the effect of different parameter values on the optimization result, as mentioned in the discussion under Section 3.1.3, and selecting a reasonable value based on the trade-off between exploration and exploitation.
Comment 3
There are some minor typos in the manuscript. The manuscript should be carefully re-examined and these errors should be corrected.
Response
The manuscript has been proofread and all minor errors have duly been rectified.
Comment 4
In Line 220, you should correct the interval as “[0, 1]”.
Response
The interval has been rectified accordingly.
Comment 5
In Line 226, you should correct the interval as “[0.4, 0.9]”. Also, why did you choose this range for the inertia weight? Please define it clearly.
Response
The explanation as to why the inertia weight was chosen has accordingly been included in the manuscript. This can be found in lines 278 – 289.
Comment 6
What is the objective function used in the study? Please define it as an equation.
Response
The objective function used in the study has been included in the manuscript. This can be found in lines 436 – 437.
Comment 7
The first letters of the names of tables and figures in the text should be capitalized like Table 1 etc.
Response
The first letters in the tables and figures have been corrected accordingly, as you suggested.
Comment 8
Please specify the limitation(s) of the study in the paper.
Response
The limitations of the study have been included in the manuscript. This can be found in 4.6, lines 581 – 587.
Comment 9
In Table 2, you have presented MSE, MAPE, and R^2 values for 10 models for PSO-BP neural network. What are these 10 models? What are the differences of these 10 models? Please specify clearly.
Response
The 10 repeated optimization models were run for the PSO-BP neural network to determine the best fitness as indicated in the table 2. In each iteration, a new population was continuously generated and the optimal weights and biases of the BP neural network were determined. This was done in order to choose the model with the best fitness. The model with the lowest global best fitness, in this case model 5, was chosen for further analysis.
Comment 10
There is no detailed information about GA-BP neural network model used in the study. You should explain this model clearly and give detailed information for this model like the design of the GA-BPNN model, etc.
Response
Accordingly, we have added some information about the GA-BP neural network model in the manuscript.
Comment
The similarity score should be rechecked and reduced.
Response
The plagiarism rate of the report has been reduced as requested. The plagiarism rate of the main manuscript with the list of references is 3%. We hope that this meets the requirements of your noble journal.
Author Response File: Author Response.docx
Round 2
Reviewer 3 Report
Suggestions for Authors:
I would recommend further reducing the similarity ratio and reviewing and correcting minor typos.
Here are some minor typos I've caught:
Line 352 Table (1) à Table 1
Line 376 inertia weight à Inertia weight
Line 396 Acceleration Coefficients à Acceleration coefficients
Line 430 fig 7 à Fig 7
Line 450 anefficient à an efficient
