Fully mechanized coal mining face gas laser sensor point distribution method

By combining NSGA-II with entropy weight TOPSIS, the deployment of gas laser sensors in fully mechanized coal mining faces was optimized, solving the problems of monitoring blind spots and redundant deployment. This enabled efficient and economical gas monitoring, improving safety and coverage.

CN122046435APending Publication Date: 2026-05-15ANHUI UNIV OF SCI & TECH +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202511868908.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-12-11
Publication Date
2026-05-15

AI Technical Summary

Technical Problem

Existing technologies for gas monitoring in fully mechanized coal mining faces suffer from problems such as blind spots, redundant deployments, and excessively high costs, making it difficult to achieve a balance between accuracy, economy, and environmental adaptability in complex environments.

Method used

The NSGA-II multi-objective optimization algorithm and the entropy weight TOPSIS evaluation method are adopted, combined with the FLUENT three-dimensional simulation model, to construct a method for the placement of gas laser sensors. By discretizing the working surface into 294 grid units, the sensor layout is optimized to maximize methane concentration, minimize dust interference, and minimize cost.

Benefits of technology

It improved the monitoring coverage of key areas by 23.5%, reduced the number of redundant sensors by 13.6%, significantly enhanced the ability to prevent and control gas disasters, shortened the response time of gas over-limit alarms to less than 3 seconds, and increased the duration of safe underground operations by 15%.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122046435A_ABST
    Figure CN122046435A_ABST
Patent Text Reader

Abstract

The invention discloses a fully mechanized coal mining face gas laser sensor distribution method, which relates to the technical field of gas laser sensors, and comprises the following implementation steps: constructing a three-dimensional simulation model, constructing the three-dimensional simulation model through FLUENT, carrying out field measurement on site, and carrying out gas laser sensor distribution. Analyzing methane and dust concentration distribution rules under different wind speeds, humidity and spatial positions, analyzing methane and dust concentration distribution rules under different wind speeds, humidity and spatial positions, and further determining dynamic diffusion characteristics of high-risk areas such as an upper corner and the like; discretizing a working face, discretizing the working face into 294 grid units, introducing multiple constraint conditions such as wind speed, humidity and concentration variance (less than or equal to 0.02), and constructing a multi-objective optimization model; and generating a non-dominated solution set, comprehensively evaluating and screening an optimal point distribution scheme through entropy weight TOPSIS, and generating the non-dominated solution set.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of gas laser sensor technology, and more specifically, to a method for deploying gas laser sensors in a fully mechanized mining face. Background Technology

[0002] Real-time monitoring of methane (primarily a gas) concentration in fully mechanized coal mining faces is crucial for preventing gas explosions and ensuring safe underground production. With increasing mining depth and mechanization, the complexity of the working face environment has significantly increased. Methane diffusion is influenced by a combination of factors, including ventilation conditions, equipment layout, and dust interference. Traditional point-based sensor deployment methods often result in delayed warnings or excessive costs due to monitoring blind spots and redundant deployments. Especially under conditions of fluctuating wind speed, humidity variations, and coal dust disturbances, methane concentration exhibits significant spatiotemporal heterogeneity. Existing research often focuses on optimizing single objectives (such as coverage or cost control), making it difficult to balance the multi-dimensional requirements of monitoring accuracy, economy, and environmental adaptability. This leads to a lack of scientific decision-making basis for sensor placement in practical engineering.

[0003] In recent years, the combination of multi-objective optimization algorithms and decision evaluation methods has provided new ideas for sensor deployment optimization in complex scenarios. NSGA-II (Non-dominated sorting genetic algorithm) excels in solving high-dimensional, nonlinear optimization problems due to its superior population diversity and convergence; while TOPSIS (Topology-to-Ideal-Solution Sorting Method) effectively quantifies the trade-offs between multiple objectives through entropy weighting, providing an objective basis for optimal solution selection. However, existing research often neglects the gas diffusion characteristics of confined spaces in downhole wells and the influence of sensor path averaging monitoring mechanisms during modeling, leading to deviations between optimization results and actual monitoring needs.

[0004] Therefore, this paper proposes a method for deploying gas laser sensors in fully mechanized mining faces. Summary of the Invention

[0005] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a method for deploying gas laser sensors in a fully mechanized mining face. The present invention provides the following technical solution: A method for deploying gas laser sensors in a fully mechanized mining face includes the following implementation steps: Step 1: Construct a 3D simulation model; Specifically, a three-dimensional simulation model was constructed using FLUENT, on-site measurements were conducted, and the distribution patterns of methane and dust concentrations under different wind speeds, humidity levels, and spatial locations were analyzed to clarify the dynamic diffusion characteristics of high-risk areas such as the upper corner. Step 2, Discretize the working surface; The working face was discretized into 294 grid cells, and multiple constraints such as wind speed, humidity and concentration variance (≤0.02) were introduced to construct a multi-objective optimization model. Step 3: Generate the non-dominated solution set; The optimal point placement scheme is selected through entropy-weighted TOPSIS comprehensive evaluation, and a non-dominated solution set is generated.

[0006] Preferably, the three-dimensional simulation model divides the working surface into multiple units, each unit having a specific concentration value, and each side of each unit is equipped with a gas laser sensor.

[0007] Preferably, the steps for constructing a multi-objective optimization model in a discretized working surface are as follows: first, the objective function is defined and quantified; then, constraint conditions are established and modeled; and finally, correlation indicators are constructed based on the feature analysis model.

[0008] Preferably, the objective function includes maximizing the average methane concentration f1, minimizing dust concentration interference f2, minimizing cost f3, and maximizing the coverage area f4; ; ; ; ; in, The methane concentration field is obtained from FLUENT simulation, and Li is the optical path length of the sensor. The mean dust concentration along the path is negatively correlated with methane monitoring signal interference. For individual sensors and installation / deployment costs; =1 indicates that grid cell j is covered by at least one sensor, and 294 is the total number of discretized grid cells on the working surface.

[0009] Preferably, the constraint modeling includes concentration stability constraints and environmental dynamic constraints; Among them, the concentration stability constraint is a limitation on the variance of path concentration. The environmental dynamic constraint is to maintain humidity at 50%~60%.

[0010] Preferably, when constructing a multi-objective optimization model, the correlation index is constructed based on the feature analysis model. The characteristics of the feature analysis model include efficiency, interference, and economy.

[0011] Preferably, the interference is characterized by minimizing the dust concentration interference of the objective function. According to the feature analysis model, the interference of dust concentration is quantified by combining the weighted average of the path dust concentration with its weights. The weight corresponding to the dust concentration is calculated as a2, and the weight corresponding to the average methane concentration is a1. Therefore, the interference index I can be defined as: In the formula, This represents the average dust concentration along the sensor's optical path.

[0012] Preferably, based on the feature analysis model, the economic efficiency of deployment costs can be quantified by a linear combination of the number of sensors and the cost per unit, with the weight corresponding to the cost being a3. Therefore, the economic efficiency index E is defined as: In the formula, The installation and maintenance of the i-th sensor.

[0013] Preferably, step 3 specifically includes the following steps: Step 301 uses the NSGA-II algorithm to solve the multi-objective model, generating a Pareto solution set, and the decision matrix composed of the Pareto solution set. Where n represents the number of solution sets, n=100. The following formulas are used to reverse and forward process f1, f2, f3, and f4 respectively, to obtain the matrix. ; ; ; Secondly, using the following formula to Standardization is performed to eliminate confusion and uncertainty caused by differences in the units of measurement between different indicators. By doing so, we can ensure that each indicator has the same magnitude throughout the research framework, thus obtaining a standardized matrix. , ; Step 302: Calculate the weights and use the following formula to calculate the information entropy values ​​of the four optimization objectives. Objective entropy weight Comprehensive entropy weight The overall entropy weight of each indicator; ; Step 303 calculates the Euclidean distance and outputs the total score. The comprehensive score Scr+ for each solution is calculated using the formula; the larger the Scr+ value, the better the result. The constructed optimization model is optimized using the NSGA-II algorithm, and the best solution set is obtained by evaluating its merits using the TOPSI method.

[0014] The technical effects and advantages of this invention are as follows: To address the issues of insufficient real-time monitoring range of methane concentration in fully mechanized mining faces and the lack of scientific basis for sensor placement, this invention proposes a collaborative optimization model for gas laser sensor placement that integrates the NSGA-II multi-objective optimization algorithm and the entropy weight TOPSIS evaluation method. First, a three-dimensional simulation model of the working face is constructed using FLUENT software. Combined with on-site measured data, the methane and dust concentrations at different roadway wind speeds, humidity levels, and locations are obtained, identifying the upper corner as a key monitoring area. Second, based on grid partitioning, the confined space of the working face is discretized into 294 units, each with corresponding data. Since the gas laser sensor measures the average methane concentration along the path, the optimization objectives are to maximize the average methane concentration, minimize the dust concentration, reduce costs, and maximize the coverage area. Furthermore, considering multiple constraints such as wind speed, humidity, and the methane concentration variance along the path not exceeding 0.02, a multi-objective optimization mathematical model is established. The NSGA-II algorithm is used to generate non-dominated solution sets. Population diversity is ensured through non-dominated sorting and crowding distance calculation. The entropy-weighted TOPSIS method is then used to comprehensively evaluate the solution sets, selecting the optimal deployment scheme that balances monitoring efficiency and economy. Simulation results show that the optimized deployment design increases the monitoring coverage of key areas by 23.5% while reducing the number of redundant sensors by 13.6%, effectively balancing monitoring accuracy and deployment costs. This method provides theoretical support for the intelligent layout of gas monitoring systems in fully mechanized mining faces, significantly improving gas disaster prevention and control capabilities. Attached Figure Description

[0015] Figure 1 This is a three-dimensional model of the working face in Embodiment 1 of the gas laser sensor placement method in a fully mechanized mining face according to the present invention.

[0016] Figure 2 This is a volume cloud map of dust concentration distribution at the working face in Example 1 of the gas laser sensor deployment method for fully mechanized mining faces according to the present invention.

[0017] Figure 3 This is a volume cloud map of dust concentration distribution at the working face in Example 1 of the gas laser sensor deployment method for fully mechanized mining faces according to the present invention.

[0018] Figure 4 This is a gas concentration distribution diagram of the working face under normal operating conditions in Example 1 of the gas laser sensor placement method for fully mechanized mining faces according to the present invention.

[0019] Figure 5 This is a flowchart of the optimization process of NSGA-II and the integrated entropy weight-TOPSIS method in the gas laser sensor placement method of a fully mechanized mining face according to the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] The implementation steps of this invention are as follows: Step 1: Construct a 3D simulation model; Specifically, this invention constructs a three-dimensional simulation model using FLUENT, conducts on-site measurements, and analyzes the distribution patterns of methane and dust concentrations under different wind speeds, humidity levels, and spatial locations. This analysis clarifies the dynamic diffusion characteristics of high-risk areas such as the upper corner.

[0022] Step 2, Discretize the working surface; To maximize the average methane concentration along the path, minimize dust interference, minimize deployment costs, and maximize coverage area, this invention discretizes the working surface into 294 grid cells and introduces multiple constraints such as wind speed, humidity, and concentration variance (≤0.02) to construct a multi-objective optimization model.

[0023] Step 3: Generate the non-dominated solution set; The optimal point placement scheme is selected through entropy-weighted TOPSIS comprehensive evaluation, and a non-dominated solution set is generated.

[0024] Example This embodiment is implemented following these steps: Step 101: Establish the geometric model and set the basic parameters.

[0025] Reference Figure 1 This embodiment takes the fully mechanized mining face of Zhongmei Xinji Liuzhuang Mine as the research object. Using SolidWorks software, a simplified model is established with the coal face roadway measuring 75m × 9.2m × 5.8m and the intake and return air roadways measuring 50m × 6.1m × 4.2m. Its construction is as follows: Figure 1As shown in the figure, this invention first constructs a three-dimensional simulation model using FLUENT, and analyzes the distribution patterns of methane and dust concentrations under different wind speeds, humidity levels, and spatial locations based on field measurement data, clarifying the dynamic diffusion characteristics of high-risk areas such as the upper corner. Secondly, the working face is discretized into 294 grid cells, with the optimization objectives of maximizing path-average methane concentration, minimizing dust interference, minimizing deployment costs, and maximizing coverage area. Multiple constraints, including wind speed, humidity, and concentration variance (≤0.02), are introduced to construct a multi-objective optimization model. Finally, NSGA-II is used to generate a non-dominated solution set, and the optimal deployment scheme is selected through entropy-weighted TOPSIS comprehensive evaluation. Simulation results show that this method significantly reduces the number of redundant sensors while improving the monitoring coverage of key areas, providing a theoretical basis and technical support for the intelligent deployment and disaster prevention of gas monitoring systems in fully mechanized mining faces.

[0026] The main parameters in Fluent are shown in the table below:

[0027] Step 102: Concentration distribution of dust and gas.

[0028] Reference Figure 2 and Figure 3 The cloud map showing the change in dust concentration in the tunnel under simulated conditions is as follows: Figure 2 As shown, the dust concentration around the coal mining machine drum is clearly a dust source, much higher than in other areas, reaching up to 3000 mg / m³. 3 The above shows that the dust concentration in the area behind the dust source decreases rapidly, and the dust concentration in the upper and middle parts of the air is much lower than that near the bottom plate. The working surface is divided into the following sections: Figure 3 The diagram shows 294 units, each with a specific concentration value, and each unit has a gas laser sensor arranged on each side.

[0029] Reference Figure 4 The physical model established based on measured parameters of the working face is divided using a structured mesh. It is assumed that the gas within the working face region consists of air and methane, and that air and methane are considered incompressible ideal gases with no chemical reactions, only convection and diffusion coupling processes. All outer boundaries except for the intake and return airways and the coal face are defined as walls, assuming no material exchange or conditional reactions with the outside environment. The gas concentration diagram of the working face region model is shown below. Figure 4 As shown.

[0030] Step 201: Construct a multi-objective optimization model.

[0031] The establishment of the objective function based on the NSGA-II algorithm and the TOPSIS evaluation framework needs to be carried out from three aspects: optimization objective quantification, constraint modeling, and multi-objective collaborative mechanism. Based on the model establishment, this model can be identified as follows: the path average concentration obtained by the gas laser sensor through optical path scanning can effectively characterize the methane concentration distribution within the discrete grid cell; the impact of dust concentration on the methane monitoring accuracy of the laser sensor can be quantified, and minimizing dust interference in the objective function can effectively reduce its interference with the monitoring signal; the roadway wind speed and humidity change stably in the short term without local unevenness; the path concentration variance limit can balance monitoring stability and accuracy, avoiding false alarms or missed alarms caused by local concentration fluctuations; the optimized sensor layout scheme meets the equipment installation spacing, power supply and communication requirements, and does not conflict with the layout of fully mechanized mining equipment (such as hydraulic supports and scraper conveyors); the sensor deployment cost only considers the unit price of a single sensor and the installation and deployment cost, without considering other costs caused by external factors.

[0032] The specific steps are as follows: ① Definition and quantization of the objective function Based on the needs of gas monitoring and the requirements of engineering economy, the following four optimization objectives are established: maximizing the average methane concentration (f1), minimizing dust concentration interference (f2), minimizing cost (f3), and maximizing the coverage area (f4). ; ; ;

[0033] in, The methane concentration field is obtained from FLUENT simulation, and Li is the optical path length of the sensor. The mean dust concentration along the path is negatively correlated with methane monitoring signal interference. For individual sensors and installation / deployment costs; =1 indicates that grid cell j is covered by at least one sensor, and 294 is the total number of discretized grid cells on the working surface.

[0034] ②Constraint modeling The following constraints apply: Concentration stability constraints, limiting the variance of path concentration. This ensures the stability of monitoring results and avoids false alarms caused by local concentration fluctuations; environmental dynamic constraints, humidity threshold (e.g., 50%~60%), where H represents temperature.

[0035] Step 203: Construct correlation indicators based on feature analysis model.

[0036] Let the original data sample set of regional gas monitoring indicators be... Where n and m are the number of monitoring points and the number of monitoring indicators, respectively, and the variables are... Let be the value taken on the i-th sample. The correlation between samples. Defined as a linear combination of variables ,in , ... Let be the weights of the m variables; y is a measure of the similarity between samples (sample association degree). The relationship between the sample association degree and the variables can be expressed as:

[0037] To find the weights of each variable (j = 1, 2, ..., m), to maximize the correlation of known samples, we define Q( Therefore, the relationship between the sample's correlation and the variable can be summarized in matrix form as Q:

[0038] The problem of the variable weight vector a is reduced to solving a system of equations;

[0039] The solution 'a' that maximizes the above system of equations using the Lagrange factorization method should satisfy the following conditions:

[0040] When λ = max When the known sample association degree is maximized, λ is a matrix. The largest eigenvalue 'a' is the eigenvector.

[0041] The key to optimizing the layout of coal mine safety monitoring points is determining the correlation degree of the samples. The magnitude of the correlation degree serves as the basis for classifying monitoring points and also reflects the environmental quality of the gas monitoring area. Based on the calculation method of the weight vector 'a' of the above variables, the weights of each variable can be calculated. Then, according to the linear equation system: The connectivity of each monitoring point is calculated in this way.

[0042] The core idea of ​​multi-objective optimization is that the functions established need to be related to each other. This invention uses efficiency, interference, and economy as the correlation models for research. Based on the above objective functions, the average methane concentration and coverage area can be used for feature analysis to obtain their respective weights to form a new function to represent efficiency. Dust concentration represents interference, and sensor and installation and deployment costs are used as economy. Thus, three major correlation models are obtained. Through multi-objective optimization, the goal of achieving high efficiency while low interference and strong economy can be achieved.

[0043] Since there is specific data on the gas concentration and dust concentration in the 294 grids, the connectivity of the monitoring points is a function obtained by multiplying the gas concentration and dust concentration by their respective weights and then summing them. This function is used to determine the efficiency of the monitoring point. The weight corresponding to the average methane concentration is calculated as a1 through the feature analysis model.

[0044] Based on the feature analysis model, its linear equation system is the sum of the degree of connection, Y; .

[0045] The interference is characterized by minimizing the dust concentration interference in the objective function. According to the feature analysis model, the interference of dust concentration can be quantified by combining the weighted mean and weights of the path dust concentration. The weights corresponding to the dust concentration are calculated as a2 through the feature analysis model. Therefore, the interference index I can be defined as: In the formula, This represents the average dust concentration along the sensor's optical path. By minimizing the interference index I, the masking effect of dust on the methane concentration monitoring signal can be effectively suppressed, thus improving the measurement accuracy of the laser sensor.

[0046] Economic efficiency is characterized by minimizing deployment costs as the objective function. According to the feature analysis model, the economic efficiency of deployment costs can be quantified by a linear combination of the number of sensors and the cost per sensor. Assuming the weight corresponding to the cost is a3 calculated through the feature analysis model, the economic efficiency index E is defined as: In the formula, Let E be the installation and maintenance cost of the i-th sensor. By minimizing the economic index E, the number of redundant sensors can be significantly reduced while meeting the constraints, achieving a cost-effective layout scheme.

[0047] Using efficiency Y, disturbance I, and economic efficiency E as the core indicators of the correlation model, a multi-objective collaborative optimization framework is constructed: Efficiency dominates coverage capability by maximizing Y to ensure methane concentration monitoring coverage and path-average concentration accuracy in key areas (such as the upper corner); interference constraints constrain environmental adaptability by minimizing I to reduce dust concentration interference on sensor optical path signals and enhance monitoring robustness under dynamic operating conditions; economy balances deployment costs by minimizing E to optimize the number and location of sensors and avoid resource waste caused by over-deployment.

[0048] Step 3, NSGA-Ⅱ algorithm and integrated entropy weight-TOPSIS.

[0049] Finally, the multi-objective model is solved using the NSGA-II algorithm to generate a Pareto solution set. Then, the solution set is comprehensively sorted using the entropy-weighted TOPSIS method to select the optimal point placement scheme.

[0050] The NSGA-II algorithm, proposed by Srinivas and Deb in 2000, is a second-generation algorithm with lower complexity and employing a non-dominated sorting method, based on NSGA. It boasts higher running efficiency. The non-dominated solution set constraint is shown in the equation. In the formula, O is the non-dominated set, M is the population, Q is the position of the offspring in space, and Xn is an arbitrary factor. The Pareto solution set is obtained by solving the problem using the NSGA-II algorithm.

[0051] TOPSIS, also known as the optimal solution distance method, is a data-driven approach that accurately reflects the differences in quality between evaluation solutions. It calculates the relative distances between all solution sets and the correct solution and ranks all solution sets according to these distances. Entropy weighting, on the other hand, focuses on calculating weights. Both methods lack a complete evaluation process and cannot fully reflect the quality of the evaluation objective. Therefore, this paper combines TOPSIS and entropy weighting to construct an evaluation method that addresses these shortcomings. The overall flowchart is as follows: Figure 5 As shown.

[0052] Step 301: Decision Matrix and Standardization. The decision matrix is ​​composed of the Pareto solution set obtained in the previous steps. Where n represents the number of solution sets, n=100; m represents the evaluation index, m=3. The following formulas are used to reverse and forward process f1, f2, f3, and f4 respectively, to obtain the matrix. ; ; Secondly, use the following formula to... Standardization is performed to eliminate confusion and uncertainty caused by differences in the units of measurement between different indicators. By doing so, we can ensure that each indicator has the same magnitude throughout the research framework, thus obtaining a standardized matrix. ;; .

[0053] Step 302: Weight Calculation. Subjective weighting may be highly subjective and arbitrary due to differing expert opinions, failing to fully consider the information in the original data; objective weighting relies solely on the data results, lacking subjective judgment. Both objective and subjective weighting alone can lead to unreasonable weight coefficients. Therefore, subjective and objective entropy weights are combined for comprehensive evaluation, and the information entropy values ​​of the four optimization objectives are calculated using the following formulas. Objective entropy weight Comprehensive entropy weight The overall entropy weight of each indicator; .

[0054] Step 303 calculates the Euclidean distance and outputs the total score. Smax+ and Smin- correspond to the maximum and minimum values ​​of each column in matrix S, respectively. The distances Dis and Dis between each solution and Smax+ and Smin- are calculated using formulas and , respectively. The comprehensive score Scr+ of each solution is calculated using formula and rearranged. The larger the value of Scr+, the better the result. .

[0055] The constructed optimization model is optimized using the NSGA-II algorithm, and the best solution set is obtained by evaluating its merits using the TOPSI method.

[0056] This invention proposes a collaborative optimization model for sensor deployment that integrates NSGA-II and entropy-weighted TOPSIS, and for the first time embeds the laser sensor path averaging monitoring mechanism into a multi-objective optimization framework. FLUENT simulations verify that the monitoring coverage rate in high-risk areas such as the upper corner is increased by 23.5%, and the number of redundant sensors is reduced by 18.6%, providing a high-precision, low-cost deployment solution for dynamic gas control. The introduction of a concentration variance constraint (≤0.02) effectively suppresses monitoring errors caused by local fluctuations, achieving a TOPSIS comprehensive score of 0.862, significantly outperforming traditional subjective weighting methods.

[0057] The method was piloted at Zhongmei Xinji Liuzhuang Mine, reducing the gas over-limit alarm response time to less than 3 seconds and increasing the safe working time underground by 15%. The model supports dynamic parameter adjustment and can be adapted to fully mechanized mining faces with different geological conditions and mining intensities.

[0058] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0059] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for deploying gas laser sensors in a fully mechanized mining face, characterized in that, The implementation steps include the following: Step 1: Construct a 3D simulation model; A three-dimensional simulation model was constructed using FLUENT, and on-site measurements were conducted. The distribution patterns of methane and dust concentrations under different wind speeds, humidity levels, and spatial locations were analyzed, thereby clarifying the dynamic diffusion characteristics of high-risk areas such as the upper corner. Step 2, Discretize the working surface; The working face was discretized into 294 grid cells, and multiple constraints such as wind speed, humidity and concentration variance (≤0.02) were introduced to construct a multi-objective optimization model. Step 3: Generate the non-dominated solution set; The optimal point placement scheme is selected through entropy-weighted TOPSIS comprehensive evaluation, and a non-dominated solution set is generated.

2. The method for deploying gas laser sensors in a fully mechanized mining face according to claim 1, characterized in that, The three-dimensional simulation model divides the working surface into multiple units, each with a specific concentration value, and each unit has a gas laser sensor arranged on each side.

3. The method for deploying gas laser sensors in a fully mechanized mining face according to claim 1, characterized in that, The steps for constructing a multi-objective optimization model in a discretized working surface are as follows: first, define and quantify the objective function; then, establish constraint conditions for modeling; and finally, construct correlation indicators based on the feature analysis model.

4. The method for deploying gas laser sensors in a fully mechanized mining face according to claim 3, characterized in that, The objective functions include maximizing the average methane concentration (f1), minimizing dust concentration interference (f2), minimizing cost (f3), and maximizing coverage area (f4). in, The methane concentration field is obtained from FLUENT simulation, and Li is the optical path length of the sensor. The mean dust concentration along the path is negatively correlated with methane monitoring signal interference. For individual sensors and installation / deployment costs; =1 indicates that grid cell j is covered by at least one sensor, and 294 is the total number of discretized grid cells on the working surface.

5. The method for deploying gas laser sensors in a fully mechanized mining face according to claim 3, characterized in that, Constraint modeling includes concentration stability constraints and environmental dynamic constraints; Among them, the concentration stability constraint is a limitation on the variance of path concentration. The environmental dynamic constraint is to maintain humidity at 50%~60%.

6. The method for deploying gas laser sensors in a fully mechanized mining face according to claim 3, characterized in that, When constructing a multi-objective optimization model, correlation indicators are built based on the feature analysis model. The characteristics of the feature analysis model include efficiency, interference, and economy.

7. A method for deploying gas laser sensors in a fully mechanized mining face according to claim 6, characterized in that, The interference is characterized by minimizing the dust concentration interference in the objective function. According to the feature analysis model, the interference of dust concentration is quantified by combining the weighted average of the path dust concentration with its weights. The weight corresponding to the dust concentration is calculated as a2, and the weight corresponding to the average methane concentration is a1. Therefore, the interference index I can be defined as: In the formula, This represents the average dust concentration along the sensor's optical path.

8. The method for deploying gas laser sensors in a fully mechanized mining face according to claim 1, characterized in that, According to the feature analysis model, the economic efficiency of deployment costs can be quantified by a linear combination of the number of sensors and the cost per sensor, with the weight corresponding to the cost being a3. Therefore, the economic efficiency index E is defined as: In the formula, Let be the installation and maintenance cost of the i-th sensor.

9. A method for deploying gas laser sensors in a fully mechanized mining face according to claim 1, characterized in that, Step 3 specifically includes the following steps: Step 301 uses the NSGA-II algorithm to solve the multi-objective model, generating a Pareto solution set, and the decision matrix composed of the Pareto solution set. Where n represents the number of solution sets, n=100. The following formulas are used to reverse and forward process f1, f2, f3, and f4 respectively, to obtain the matrix. ; Secondly, using the following formula to Standardization is performed to eliminate confusion and uncertainty caused by differences in the units of measurement between different indicators. By doing so, we can ensure that each indicator has the same magnitude throughout the research framework, thus obtaining a standardized matrix. , ; Step 302: Calculate the weights and use the following formula to calculate the information entropy values ​​of the four optimization objectives. Objective entropy weight Comprehensive entropy weight The overall entropy weight of each indicator; Step 303 calculates the Euclidean distance and outputs the total score. The comprehensive score Scr+ for each solution is calculated using the formula; the larger the Scr+ value, the better the result. The constructed optimization model is optimized using the NSGA-II algorithm, and the best solution set is obtained by evaluating its merits using the TOPSI method.