Research on Section Coal Pillar Deformation Prediction Based on Fiber Optic Sensing Monitoring and Machine Learning Algorithms
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsSUMMARY
In this article the ensemble empirical modal decomposition (EEMD) is introduced for primary noise reduction and signal reconstruction by the threshold determination (DE) algorithm is presented. The singular matrix decomposition (SVD) is adapted for secondary noise reduction, and finally, a machine learning algorithm is combined with the noise reduction algorithm for the prediction of fiber grating strain signals of coal pillar in a zone, and DBO-LSTM-BP is constructed as the prediction model.
The article is well organized and can be accepted by minor corrections. However, there are several concepts which are unclear or need further explanations.
COMMENTS
1. In your article you proposed fiber-optic monitoring method for real life monitoring of coal pillar deformation. Dear authors, what is your opinion about 3D laser scanner technology, which also could be used for monitoring of coal pillar deformation? What are plus and cons of this method?
2. One of the biggest problem of monitoring the structures such as coal pillars is anisotropy of the material. Also, the grain sizes are different. Dear authors, would like to know what is the variation of the min and max sizes of the grains? Because it‘s very important for the signal reconstruction and denoising.
3. Line 183: Dear authors, you made your experiments from July 9 to July 29. It‘s around 1 month. Is it possible to use your technique for a long lasting measurements and monitoring, for example 1 year or longer?
4. In general, the quality of imaging is defined by the fact how clearly the reflections of defects appear above the background noise. The following parameters could be proposed for assessment of the optical image quality: conventional signal to noise ratio, signal to mean noise ratio, signal to mean squared noise ratio, contrast to noise ratio. Why did you prefer SNR exactly?
5. Would like to know what is your future research directions and where exactly would you like to use your proposed methods and techniques?
Finally, I would like to say that it was a big pleasure to read this article
Author Response
I would like to express my sincere gratitude for the invaluable revisions you have made to this article, especially given the time constraints. In response to the your suggestions, the authors have addressed each comment systematically in the main text, with corresponding corrections highlighted in red font. Below is our detailed response to the comments provided by you:
Comment 1: In your article you proposed fiber-optic monitoring method for real life monitoring of coal pillar deformation. Dear authors, what is your opinion about 3D laser scanner technology, which also could be used for monitoring of coal pillar deformation? What are plus and cons of this method?
Response: Thank you for your comments. Regarding the view that 3D laser scanning technology is used for coal pillar deformation monitoring, I think it is also a very effective technology, especially because it has great advantages in high-precision measurement and three-dimensional spatial deformation monitoring. However, this technology is mainly suitable for the monitoring of surface deformation, which cannot reflect the strain inside the coal pillar, and has high environmental requirements, such as dust and humidity in the mine, which may affect the accuracy. Therefore, although 3D laser scanning technology has excellent performance in surface monitoring, it is different from 3D laser scanning technology, distributed optical fiber monitoring has high precision, long-term real-time and anti-interference, which is conducive to the internal monitoring of underground coal pillar deformation, and optical fiber monitoring is more suitable for internal strain monitoring, and the two can complement each other in practical applications.
Comment 2: One of the biggest problem of monitoring the structures such as coal pillars is anisotropy of the material. Also, the grain sizes are different. Dear authors, would like to know what is the variation of the min and max sizes of the grains? Because it‘s very important for the signal reconstruction and denoising.
Response: Thank you for your comments. Concerning the variation in the minimum and maximum particle sizes in coal pillar materials, the particle size distribution may fluctuate significantly due to the complexity and natural structure of the material. Typically, the particle sizes in coal pillars range from micron-sized fine particles to centimeter-sized coarse particles. The difference in particle sizes has a crucial impact on signal reconstruction and noise reduction. Larger particles may cause irregular signal fluctuations, while smaller particles may introduce high-frequency noise. To address these variations in particle size, we have employed methods such as ensemble empirical mode decomposition (EEMD) and singular value decomposition (SVD) to effectively mitigate the noise caused by these differences and ensure accurate signal reconstruction.
Comment 3: Line 183: Dear authors, you made your experiments from July 9 to July 29. It‘s around 1 month. Is it possible to use your technique for a long lasting measurements and monitoring, for example 1 year or longer?
Response: Thank you for your comments. we have tested it in our experiments for about 3 months, and therefore, it has potential for long-term measurement and monitoring. Fiber grating sensors have good long-term stability and resistance to environmental interference, especially in long-term strain monitoring of structures such as coal pillars, which enables continuous and reliable data acquisition. So the method can be applied for monitoring up to 1 year or more. However, long-term monitoring in practical applications also needs to consider the effects of environmental factors on the equipment, such as temperature changes, humidity, and dust in the mine. These factors may affect the long-term performance of the sensors, but with appropriate protective measures and regular maintenance, it is still possible to ensure the reliable operation of the system.
Comment 4: In general, the quality of imaging is defined by the fact how clearly the reflections of defects appear above the background noise. The following parameters could be proposed for assessment of the optical image quality: conventional signal to noise ratio, signal to mean noise ratio, signal to mean squared noise ratio, contrast to noise ratio. Why did you prefer SNR exactly?
Response: Thank you for your comments. In our study, we chose to use signal-to-noise ratio (SNR) as the primary metric for evaluating image quality for several reasons. SNR is a commonly used and effective measure of the signal strength relative to background noise, directly indicating whether the signal is prominent enough. This is particularly useful in coal pillar monitoring, where SNR can assess the contrast between defect or strain signals and noise. Compared to other metrics, such as signal-to-mean noise ratio or signal-to-mean squared noise ratio, SNR provides reliable noise evaluation results while maintaining computational efficiency. Hence, we selected SNR as the main metric to simplify the evaluation process and ensure accurate and practical results.
Comment 5: Would like to know what is your future research directions and where exactly would you like to use your proposed methods and techniques?
Response: Thank you for your comments. My future research direction will continue to focus on the field of intelligent mining, especially for real-time monitoring and intelligent analysis in complex mine environments. Specifically, I hope to further optimize the combination of optical fiber monitoring technology and machine learning algorithms, realize artificial intelligence and big data collaborative monitoring system, and better guide mine safety production. At the same time, I also hope to explore the integration of more sensing technologies, such as combining 3D laser scanning, geological radar and other means to build a monitoring system with multi-source data fusion, so as to provide comprehensive monitoring and prediction solutions for intelligent mining. These technologies can be widely used in the health monitoring of mine support structures, the prediction of rock deformation in the mining process, and the early warning system in dangerous areas, helping to achieve safe and efficient production in mining areas and promote the development of intelligent mining.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors justify the need for research by the lack of effective methods to realize real-time and accurate deformation monitoring of the internal deformation of the section coal pillars. They rightly note that traditional methods such as: numerical simulation, drilling peeping, and acoustic emission are not suitable for use in the mining face under the close coal seam group is affected by the superposition of the concentrated stress of the overlying residual diagonally intersecting coal pillar and the mining stress. The research gap has therefore been properly identified and defined. The authors also indicate the location of coal deposits in China, where the results of their research can be used. The aim of the research is therefore justified both practically and theoretically. It also fits the scope of the journal.
However, the literature review is not sufficient and requires supplementation. 20 items are definitely too few to illustrate the achievements to date in the analyzed area. The review must justify the need for research more clearly.
The ensemble empirical modal decomposition (EEMD) is introduced for primary noise reduction and signal reconstruction by the threshold determination (DE) algorithm, and then the singular matrix decomposition (SVD) is introduced for secondary noise reduction, and finally, a machine learning algorithm is combined with the noise reduction algorithm for the prediction of fiber grating strain signals of coal pillar in a zone, and DBO-LSTM-BP is constructed as the prediction model. The methodology used is trichonomic, appropriate and justified. The methodology enables replication of the performed analyses.
The results are well described and documented. The graphic material enriches the text and is well developed. The conclusions are a summary of the research. They lack: - indication of originality, - the scope of filling the research gap, - research limitations, - indications for practical use, - directions for further research and analysis. The above must be supplemented.
Author Response
I would like to express my sincere gratitude for the invaluable revisions you have made to this article, especially given the time constraints. In response to the your suggestions, the authors have addressed each comment systematically in the main text, with corresponding corrections highlighted in red font. Below is our detailed response to the comments provided by you:
Comment 1: The literature review is not sufficient and requires supplementation. 20 items are definitely too few to illustrate the achievements to date in the analyzed area. The review must justify the need for research more clearly.
Response: Thank you for your comments.
According to the experts' opinions, the synthesis part of the thesis has been improved.
The research background of the thesis, the current status of the research on deformation monitoring of coal pillars in the section, as well as the fiber optic sensing technology used in mine monitoring part of the expanded the scope of literature search, increased the number of cited articles, illustrated the achievements of the experts in the analyzed field so far, as well as the deficiencies in some aspects of the current existence of the content of this part to make additions to this part of the content of the improvement.
Comment 2: The results are well described and documented. The graphic material enriches the text and is well developed. The conclusions are a summary of the research. They lack: - indication of originality, - the scope of filling the research gap, - research limitations, - indications for practical use, - directions for further research and analysis. The above must be supplemented.
Response: Thank you for your comments.
According to the experts' opinions, we have made additional explanations for the conclusion part.
The overall originality of the article is explained, the main research content and innovativeness of the article are pointed out, and the limitations and scope of application of the study are added, as follows.
In this paper, the internal deformation of section coal pillar is monitored by fiber grating sensing technology, and a prediction model of section coal pillar deformation is constructed by combining machine learning. The model is applicable to the conditions of shallow buried close coal seam section coal pillar stability monitoring, working face mining on section coal pillar internal deformation monitoring and broken ring prediction. In addition, the model combines two noise reduction algorithms, which helps to improve the prediction accuracy of the model. The accuracy and reliability of the DBO-LSTM-BP prediction model is confirmed by comparing it with the measured data in the field. The main findings of this study are summarized as follows.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsThe article could be published now.
