A mine roadway air volume optimization adjusting method

Abstract

The application discloses a kind of mine roadway air volume optimization regulation methods, the upper limit of target wind demand branch adjustable air volume is optimization target in this method, builds underground ventilation network optimization calculation model, introduces non-differentiable accurate penalty function to implement transformation processing to constraint problem in optimization process, and determines air volume optimization regulation strategy in combination with air volume sensitivity theory;Quote the mathematical model of multi-strategy collaborative improvement whale migration optimization algorithm to optimize the air resistance value of adjustable branch, to complete the quick, accurate regulation and control of underground air volume.The present application can properly solve the catastrophic condition caused by the concentration of toxic and harmful gases such as gas in the target wind demand branch exceeding the standard, and ensure the long-term stable operation of the underground ventilation system.

Description

Technical Field

[0001] This invention relates to the field of mine ventilation, and in particular to a method for optimizing and adjusting the air volume in mine roadways. Background Technology

[0002] Mine ventilation systems perform two key functions: continuously supplying clean fresh air to underground mining areas and promptly extracting hazardous gases such as methane and carbon monoxide generated underground. They are core infrastructure for maintaining the health and safety of underground workers and strengthening the mine's safety production line. During the continuous advancement of underground mining operations, the layout of various ventilation structures and the connectivity of the ventilation network dynamically change, easily leading to situations where the actual air volume at work points fails to meet the design air supply standards. As the mining face extends forward, basic ventilation parameters such as the effective ventilation cross-section of the roadway and the roughness resistance coefficient of the surrounding rock will fluctuate dynamically, thus disrupting the overall ventilation network structure and weakening the stability and reliability of the ventilation system. If the air volume of a specific ventilation branch does not meet production requirements, or if a sudden abnormal disturbance causes a rapid increase in the required air volume for that branch, the air volume is usually redistributed by adjusting the air resistance of adjacent branches to meet on-site production needs.

[0003] With the increasing prevalence of deep mining and multi-regional simultaneous mining models, the total number of underground development roadways continues to expand, and the spatial layout of ventilation points in various mining operations is also frequently changing. These multiple factors significantly increase the difficulty of routine monitoring, control, and scheduling management of the ventilation system. If there is insufficient ventilation supply in some roadway branches, flammable and explosive gases such as methane can easily accumulate in confined or poorly ventilated areas. Once the concentration of harmful gases exceeds the safety control threshold, harmful gas spills and diffusions will occur, easily triggering mine safety accidents and creating a continuous safety hazard to the overall underground working environment.

[0004] Therefore, how to optimize the air volume allocation of branch ventilation roadways to achieve efficient response and precise control of mine ventilation has become a pressing problem that needs to be solved in current underground ventilation systems. Summary of the Invention

[0005] In order to optimize and adjust the air volume of the branch roadway, achieve rapid and accurate control of the mine air volume, and thus stabilize the ventilation system, this application provides a method for optimizing and adjusting the air volume of the mine roadway.

[0006] This application provides a method for optimizing and adjusting ventilation volume in mine roadways, employing the following technical solution: A method for optimizing and adjusting ventilation volume in mine roadways, comprising: S1: First, identify the target air demand branch that needs airflow optimization and adjustment. Calculate the airflow sensitivity matrix corresponding to the ventilation network to which the target air demand branch belongs using the airflow sensitivity calculation formula. In the airflow sensitivity matrix, sort the airflow sensitivity of each branch acting on the target air demand branch according to their numerical values. Use the resistance-increasing airflow adjustment method to adjust the airflow, selecting branches with airflow sensitivity greater than 0 as the adjustable branch set. Use regression analysis to establish a functional relationship between airflow sensitivity and roadway resistance, calculating the reasonable control range of the resistance parameter. Use a multi-strategy collaborative improvement whale migration optimization algorithm mathematical model to optimize the resistance value of the adjustable branches. S2: with To define the objective function for regulating airflow in the ventilation network, initialize population information and relevant parameters, including population size. Maximum number of iterations The variable dimension corresponding to the number of adjustable branches. The upper limit of the solution space corresponding to the adjustable branch wind resistance adjustment range and lower limit The adjustable branch's wind resistance adjustment value is The upper limit of the solution space corresponding to the adjustable branch wind resistance adjustment range and lower limit This refers to the reasonable adjustment range for wind resistance; S3: Initialize the population using the optimal point set method and randomly generate the initial population. ; S4: Substitution Calculate each individual air volume adaptability value Obtain the current maximum airflow fitness value and the location of the corresponding individual. And record it, among which The optimal solution for the air conditioning parameters represents the best air resistance adjustment value for the adjustable branch; S5: For each individual air volume adaptability value Sort the individuals and select the top 20% with the best fitness values ​​as the leaders of the population, and the bottom 80% with the worst fitness values ​​as the followers of the population. S6: After updating the current position of the population leader according to the leader's position update formula, the individual position is updated again using the golden sine algorithm strategy; the current position of the population followers is updated according to the improved follower position update formula. S7: After completing the individual position update, a two-way population evolution dynamics strategy is adopted to update the position of each whale individual; S8: Compare the airflow adaptability values ​​for each individual. Preserve the optimal airflow adaptation value and the corresponding optimal location. Determine if the current number of iterations reached is the maximum. If it is, execute S9; otherwise, return to execute S5. S9: Output the global optimal solution and the corresponding optimal airflow fitness value This yields the set of adjustable branches that can meet the airflow requirements of the branch requiring airflow, and the optimal air resistance adjustment of the adjustable branches. The optimal air resistance adjustment of the adjustable branches is the global optimal solution. This ensures that the airflow of the branch ventilation system meets the ventilation requirements.

[0007] Optionally, when using a multi-strategy collaborative improvement algorithm mathematical model for whale migration optimization to optimize the wind resistance value of adjustable branches, the wind resistance value of the branch with the highest airflow sensitivity value in the adjustable branch set is first optimized and adjusted. If the highest branch meets the airflow requirement of the branch requiring airflow, then the highest branch is the optimal adjustable branch set; if it does not meet the airflow requirement of the branch requiring airflow, then the wind resistance value of the branch with the second highest airflow sensitivity value in the adjustable branch set is adjusted; if it still does not meet the airflow requirement of the branch requiring airflow, then the wind resistance value of the branch with the third highest airflow sensitivity value in the adjustable branch set is adjusted; and so on, until the airflow requirement of the branch requiring airflow can be met, then these branches are the optimal adjustable branch set.

[0008] Optionally, in S3, the formulas for initializing the population individuals using the optimal point set method are shown in equations (14) and (15): The principle of optimal point set is based on the assumption that... The unit cube in Euclidean space There exists a set of points in the set, represented by equation (14):

[0009] Its deviation satisfies ,in, It is only with The relevant constants, at this time, It's a collection of great spots. For the optimal point, its value is ,in, It is to satisfy The smallest prime number, mapped to the initial population, can be expressed as equation (15): .

[0010] Optionally, in S6, after the population leader updates its current position according to the leader's position update formula, the individual positions are updated again using the golden sine algorithm strategy; the current positions of the population followers are updated according to the improved follower position update formula: The formula for updating the position of the population leader is Equation (11):

[0011] In the formula: This represents the current iteration number; For the first The whale in the The position at the next iteration; For the first The whale in the The position at the next iteration; and All Random numbers within the interval; For the number of leader whales, ; The formula (16) for updating the individual position again using the golden sine algorithm strategy is:

[0012] In the formula: for The random number within the interval determines the distance an individual moves during the iteration; for Random numbers within the interval determine the direction of position updates for an individual during iteration; the golden ratio. , The golden ratio It is an irrational number; The improved follower position update formula (17) is as follows:

[0013] In the formula: For the first The whale in the Fitness value at the next iteration; For the first The whale in the Fitness value at the next iteration; For interval The random number, where As shown in equation (18); 1 represents the adaptive weight, as shown in equation (19); and For individuals randomly selected from followers Two other different whales and whales The location.

[0014]

[0015]

[0016] In the formula: This represents the maximum number of iterations. Optionally, in S7, when the bidirectional population evolution dynamics strategy updates the position of each whale individual, new mutant individuals are obtained for each whale individual of the whale leader through natural mutation, as shown in Equation (20).

[0017] The whale followers are randomly divided into two groups. One group of whales will mutate around the best individual in the group, as shown in Equation (21). The other group of whales will relocate to their initial position for further exploration, as shown in Equation (22).

[0018]

[0019] In the formula: For the first The location of an individual whale after mutation; The mutation factor is shown in equation (23); and For individuals randomly selected from the leaders The other two different whales and whales Location; , ;

[0020] For each individual whale, the better individuals before and after the mutation are retained, as shown in Equation (24).

[0021]

[0022] In the formula: For the first The fitness value of each individual whale after mutation; the position of the population leader is updated according to equations (20) and (24), and the random number is determined. If the value is less than 0.5, the follower position is updated according to equations (21) and (24); otherwise, the follower position is updated according to equations (22) and (24).

[0023] In summary, this application includes at least one of the following beneficial technical effects: 1. This method takes the upper limit of adjustable air volume of the target ventilation branch as the optimization target, builds an optimization calculation model of the underground ventilation network, introduces a non-differentiable precise penalty function to transform the constraint problem in the optimization process, and determines the air volume optimization adjustment strategy by combining air volume sensitivity theory; it uses a multi-strategy collaborative improvement whale migration optimization algorithm mathematical model to optimize the wind resistance value of the adjustable branch, thereby achieving rapid and accurate control of underground air volume; this application can properly solve the disaster working conditions caused by the excessive concentration of toxic and harmful gases such as methane in the target ventilation branch, and ensure the long-term stable operation of the underground ventilation system. Attached Figure Description

[0024] Figure 1 This is a preferred point set diagram of the embodiments of this application.

[0025] Figure 2 This is a point graph of the original algorithm in the embodiments of this application.

[0026] Figure 3 This is a flowchart of the MSCWMA calculation process in the embodiment of this application.

[0027] Figure 4 This is a fitness convergence curve of the five algorithms on the F1 test function.

[0028] Figure 5 This is a fitness convergence curve of the five algorithms on the F2 test function.

[0029] Figure 6 This is a fitness convergence curve of the five algorithms on the F3 test function.

[0030] Figure 7 This is a fitness convergence curve of the five algorithms on the F4 test function.

[0031] Figure 8 This is a fitness convergence curve of the five algorithms on the F5 test function.

[0032] Figure 9 This is a fitness convergence curve of the five algorithms on the F6 test function.

[0033] Figure 10 This is a topology diagram of a mine ventilation network.

[0034] Figure 11 It is the sensitivity d 8,16 Wind resistance R 16 The trend chart.

[0035] Figure 12 It is the sensitivity d 8,4 Trend graph of wind resistance R4.

[0036] Figure 13 It is the sensitivity d 8,9Trend graph of wind resistance R9.

[0037] Figure 14 This is an iterative graph of five algorithms optimizing the airflow in branch 8. Detailed Implementation

[0038] The present application will be further described in detail below with reference to all the accompanying drawings.

[0039] This application discloses a method for optimizing and adjusting ventilation in mine roadways, including: I. Mine Ventilation Network Calculation and its Mathematical Model During mine production, when a certain ventilation branch experiences insufficient airflow or an emergency leading to increased airflow demand, the airflow to that branch is regulated by adjusting the air resistance of other branches. To address this, an optimization function targeting the airflow of the ventilation branch is established. As shown in equation (1):

[0040] In the formula: For the airflow of the target branch, m 3 / s.

[0041] During the airflow regulation process, the ventilation network follows the following balance equation, as shown in equation (2):

[0042] In the formula: For elements in the correlation matrix, Indicates branch Wind flow into the node , Indicates branch Wind outflow node , Represents a node Not a branch The endpoints; For branches Air volume; For elements in the loop matrix, Indicates branch AND loop Same direction, Indicates branch AND loop Reverse, Indicates branch Not in the loop middle; For branches Wind resistance; For loop The algebraic sum of natural wind pressure in the equation; For loop The algebraic sum of mechanical wind pressure in the equation; For branches Ventilation resistance.

[0043] Based on the established air volume objective function, combined with the basic laws of air volume flow and the constraints of air volume regulation, a basic mathematical model for optimizing air volume regulation in a mine ventilation network can be established, as shown in equation (3).

[0044]

[0045] In the formula: For branches Minimum air volume; For branches Maximum air volume; For branches Minimum value of the adjustable range of wind resistance; For branches The maximum value of the adjustable range of wind resistance; This refers to the operating air pressure of the ventilation fan; This refers to the maximum air pressure of the ventilation fan; This refers to the working efficiency of the ventilation fan.

[0046] This application employs the exact penalty function method to further optimize the model for airflow regulation in the ventilation network. The optimized nonlinear unconstrained airflow objective function... As shown in equation (4).

[0047]

[0048] In the formula: , , As a penalty factor, in The value can be taken from within the range.

[0049] II. Airflow Sensitivity This application uses air volume sensitivity to measure the degree of change in roadway air volume after being affected by air resistance. The expression for air volume sensitivity is shown in equation (5):

[0050] In the formula: For branches air volume Relative to branches wind resistance Sensitivity to changes.

[0051] For those with For a branched ventilation network, its airflow sensitivity matrix is ​​denoted as... As shown in equation (6).

[0052]

[0053] Where: the first of the sensitivity matrix Row elements represent branches Sensitivity of airflow to changes in the resistance of each branch; Column elements represent branches The sensitivity of changes in wind resistance to changes in air volume in each branch.

[0054] The airflow sensitivity matrix of the ventilation network is solved analytically, and the sensitivity is simulated by computer. Wind resistance Based on data analysis and fitting, the sensitivity was established using regression analysis. and wind resistance The data relationship between them is shown in equation (7):

[0055] In the formula: , , All of these are constants greater than zero, which can be obtained through data fitting.

[0056] The critical value of wind resistance for adjusting the branch can be obtained based on empirical formulas. As shown in equation (8):

[0057] In the formula: For branches The initial wind resistance.

[0058] Therefore, if you want to make the drag-increasing airflow regulation more sensitive and efficient, the branch... wind resistance It must be kept within a reasonable range, that is It should satisfy equation (9):

[0059] At this point, the reasonable range for adjusting the wind resistance of the adjustable branch can be determined.

[0060] III. Whale Migration Algorithm (I) Mathematical Model of Standard Whale Migration Algorithm The Whale Migration Algorithm (WMA) is a novel metaheuristic algorithm that simulates the cooperative migration process of whale groups to achieve global search and local exploitation in the solution space, thereby finding the optimal solution to the problem.

[0061] (ii) Population initialization phase In WMA, the initial population of migrating whales is randomly generated between the lower and upper limits of the search space, as shown in Equation (10).

[0062]

[0063] In the formula: For the first One initial population of individuals; For dimension size; for dimension A vector of random numbers within an interval; This is the lower bound of the search space; This represents the upper limit of the search space. Population size.

[0064] (III) Leadership Renewal Phase In migratory whale groups, more experienced individuals are chosen as leaders to identify and select the optimal route to the destination, guiding and directing the group towards it. In WMA, the top 20% of migratory whales with higher fitness values ​​are selected as leaders, and the leader position update formula is shown in equation (11):

[0065] In the formula: This represents the current iteration number; For the first The whale in the The position at the next iteration; For the first The whale in the The position at the next iteration; and All Random numbers within the interval; For the number of leader whales, .

[0066] (iv) Follower Update Phase During whale migration, less experienced whales are guided by more experienced whales to help them reach their destination more easily. In WMA, the bottom 80% of migrating whales with the worst individual fitness values ​​are considered followers, and the position update formula for these followers is shown in equation (12).

[0067] In the formula: For the present The average position of the leaders is shown in Equation (13); For the first The whale in the The position at the next iteration; This represents the position of the current optimal individual.

[0068] .

[0069] IV. Multi-strategy Collaborative Improvement Algorithm for Whale Migration To enhance the optimization capability of WMA, this application proposes an MSCWMA: First, a set of optimal points is introduced to optimize the initial population distribution and improve population diversity; second, a golden sine algorithm update strategy is adopted after the leader update phase to accelerate the optimization speed; then, a global exploration strategy of the slime mold optimization algorithm (SMA), which has a strong global optimization capability, is incorporated into the follower update phase to expand the search range and enhance the global search capability of WMA; finally, a bidirectional population evolution dynamics strategy is adopted in the individual position update to improve population diversity and the algorithm's search capability.

[0070] (a) Optimal Point Set Population Initialization Strategy In standard WMA, the initial population is set up randomly, which may result in over-clustering or over-dispersion of the population, affecting the global search capability of WMA. To make the initial population distribution more even, the optimal point set method is used to improve population initialization. This not only increases population diversity but also improves global search capability during the initialization phase, helping the algorithm escape local optima. The principle of the optimal point set is based on the assumption that... The unit cube in Euclidean space There exists a set of points in the set, represented by equation (14):

[0071] Its deviation satisfies ,in, It is only with The relevant constants, at this time, It's a collection of great spots. For the optimal point, its value is ,in, It is to satisfy The smallest prime number, mapped to the initial population, can be expressed as shown in equation (15):

[0072] Using MATLAB, we simulated the generation of randomly distributed two-dimensional initial population point maps using both the optimal point set algorithm and the original algorithm. Both algorithms used a population size of 100 and a dimension of 2. The two-dimensional point maps generated by the two initialization methods are shown below. Figure 1 and Figure 2 As shown.

[0073] Depend on Figure 1 and Figure 2 It can be observed that the optimal point set method can indeed achieve population homogenization and diversification in population initialization, and is helpful in escaping local optima.

[0074] (II) Golden Sine Algorithm Update Strategy During the whale leader update phase, the whale moves randomly around the main linear direction, which can easily lead to local maxima. Therefore, this application employs the golden sine algorithm after the leader update phase to expand the search range and make the solution process more flexible. The golden sine algorithm has the characteristics of fast convergence, good robustness, and simple parameter tuning. Its position update formula is shown in equation (16):

[0075] In the formula: for The random number within the interval determines the distance an individual moves during the iteration; for Random numbers within the interval determine the direction of position updates for an individual during iteration; the golden ratio. , The golden ratio It is an irrational number.

[0076] When updating positions using the golden sine algorithm, individual whales interact with the optimal individual during iterations, measuring the position difference and utilizing the golden ratio. , Optimize the search area and use parameters , By controlling the movement distance and direction during iteration, the ergodicity of the solution space is enhanced, allowing individuals to steadily approach the optimal value.

[0077] (III) Integrating SMA Strategy In the follower position update process of WMA, followers move towards the current optimal individual position. Although this behavior can promote rapid population aggregation and improve the convergence speed, it also reduces population diversity and increases the probability of the population getting trapped in local optima. To solve the above problems, this application incorporates the global exploration strategy of SMA, which has a strong global optimization capability, into the follower update stage to expand the search range and enhance the global search capability of WMA. SMA shows good exploration performance in complex multimodal function optimization problems and has a strong exploration capability. The improved follower position update formula is shown in Equation (17):

[0078] In the formula: For the first The whale in the Fitness value at the next iteration; For the first The whale in the Fitness value at the next iteration; For interval The random number, where As shown in equation (18); The adaptive weights are shown in equation (19); and For individuals randomly selected from followers Two other different whales and whales The location.

[0079]

[0080]

[0081] In the formula: This represents the maximum number of iterations.

[0082] (iv) Two-way population evolution dynamics strategy To prevent the algorithm from getting stuck in local optima in the later stages of iteration, a two-way population evolution dynamics strategy is introduced in the population individual position update to improve the population diversity and optimization capability of the standard WMA. For each whale individual of the whale leader, new mutant individuals are obtained through natural mutation, as shown in Equation (20); for whale followers, they are randomly divided into two groups. One group of whale individuals will mutate around the best individual in the group, as shown in Equation (21), and the other group of whale individuals will relocate near their initial position for further exploration, as shown in Equation (22).

[0083]

[0084]

[0085]

[0086] In the formula: For the first The location of an individual whale after mutation; The mutation factor is shown in equation (23); and For individuals randomly selected from the leaders The other two different whales and whales Location; , .

[0087]

[0088] For each individual whale, the better individuals before and after the mutation are retained, as shown in Equation (24).

[0089]

[0090] In the formula: For the first The fitness value of an individual whale after mutation.

[0091] V. Airflow Optimization Solution Based on MSCWMA Algorithm Based on the above explanation of the principles and mathematical model of MSCWMA, the solution process for MSCWMA can be obtained as follows: Figure 3 As shown.

[0092] This application adopts an improved WMA to form a mine ventilation volume optimization and regulation method based on MSCWMA. The following are the steps for applying MSCWMA to solve the mine ventilation volume optimization problem: S1: First, identify the target air demand branch that needs airflow optimization and adjustment. Calculate the airflow sensitivity matrix corresponding to the ventilation network to which the target air demand branch belongs using the airflow sensitivity calculation formula. In the airflow sensitivity matrix, sort the airflow sensitivity of each branch acting on the target air demand branch according to their numerical values. Use the resistance-increasing airflow adjustment method to adjust the airflow, selecting branches with airflow sensitivity greater than 0 as the adjustable branch set. Use regression analysis to establish a functional relationship between airflow sensitivity and roadway resistance, calculating the reasonable control range of the resistance parameter. Use a multi-strategy collaborative improvement whale migration optimization algorithm mathematical model to optimize the resistance value of the adjustable branches. S2: with To define the objective function for regulating airflow in the ventilation network, initialize population information and relevant parameters, including population size. Maximum number of iterations The variable dimension corresponding to the number of adjustable branches. The upper limit of the solution space corresponding to the adjustable branch wind resistance adjustment range and lower limit The adjustable branch's wind resistance adjustment value is The upper limit of the solution space corresponding to the adjustable branch wind resistance adjustment range and lower limit This refers to the reasonable adjustment range for wind resistance; S3: Initialize the population using the optimal point set method and randomly generate the initial population. ; S4: Substitution Calculate each individual air volume adaptability value Obtain the current maximum airflow fitness value and the location of the corresponding individual. And record it, among which The optimal solution for the air conditioning parameters represents the best air resistance adjustment value for the adjustable branch; S5: For each individual air volume adaptability value Sort the individuals and select the top 20% with the best fitness values ​​as the leaders of the population, and the bottom 80% with the worst fitness values ​​as the followers of the population. S6: After updating the current position of the population leader according to the leader's position update formula, the individual position is updated again using the golden sine algorithm strategy; the current position of the population followers is updated according to the improved follower position update formula. S7: After completing the individual position update, a two-way population evolution dynamics strategy is adopted to update the position of each whale individual; S8: Compare the airflow adaptability values ​​for each individual. Preserve the optimal airflow adaptation value and the corresponding optimal location. Determine if the current number of iterations reached is the maximum. If it is, execute S9; otherwise, return to execute S5. S9: Output the global optimal solution and the corresponding optimal airflow fitness value This yields the set of adjustable branches that can meet the airflow requirements of the branch requiring airflow, and the optimal air resistance adjustment of the adjustable branches. The optimal air resistance adjustment of the adjustable branches is the global optimal solution. This ensures that the airflow of the branch ventilation system meets the ventilation requirements.

[0093] When optimizing the wind resistance value of adjustable branches using the multi-strategy collaborative improvement whale migration algorithm mathematical model, the wind resistance value of the branch with the highest airflow sensitivity value in the adjustable branch set is first optimized and adjusted. If the highest branch meets the airflow requirement of the branch requiring airflow, then the highest branch is the optimal adjustable branch set; if it does not meet the airflow requirement of the branch requiring airflow, then the wind resistance value of the branch with the second highest airflow sensitivity value in the adjustable branch set is adjusted; if it still does not meet the airflow requirement of the branch requiring airflow, then the wind resistance value of the branch with the third highest airflow sensitivity value in the adjustable branch set is adjusted; and so on, until the airflow requirement of the branch requiring airflow can be met, then these branches are the optimal adjustable branch set.

[0094] VI. Performance Testing and Result Analysis of the MSCWMA Algorithm To verify the overall optimization performance of MSCWMA, this application selected four algorithms with good optimization performance, including the standard WMA, ALA, WHO, and WOA, for comparison. The performance of MSCWMA was evaluated through six benchmark test functions, which are shown in Table 1.

[0095] Table 1 Benchmark Test Functions

[0096] Using MATLAB, five algorithms were tested on the six benchmark functions listed in Table 1. The parameter settings for the five algorithms are as follows: population size Maximum number of iterations Considering the randomness of a single algorithm run, all algorithms were run 20 times in the experiment, and the optimization results are shown in Table 2.

[0097] Table 2. Optimization results of the five algorithms on the benchmark function.

[0098] To more intuitively analyze the convergence speed and ability to avoid local optima of the algorithms, the fitness convergence curves of the five algorithms on six test functions are shown below. Figures 4 to 9 As shown.

[0099] From Table 2, Figures 4 to 9 It can be seen that MSCWMA performs well in benchmark tests. Its solution accuracy is superior to ALA, WHO, WOA, and WMA, while ALA, WHO, WOA, and WMA only perform better in functions. The solution accuracy is the same as that of MSCWMA in the function. Among them, MSCWMA is the first to achieve the global optimum, followed by ALA, WHO, WOA and WMA. Therefore, the convergence speed of MSCWMA is also better than the other four algorithms.

[0100] In summary, compared with ALA, WHO, WOA, and WMA, MSCWMA has the fastest convergence speed and the highest convergence accuracy, and has better overall stability and global search capability. The other four algorithms are prone to getting trapped in local optima and have slower convergence speeds, further verifying the superior optimization performance of MSCWMA.

[0101] VII. Application Research of MSCWMA Algorithm in Mine Airflow Optimization and Regulation (I) Construction of the experimental mine and calculation of the air volume of its ventilation system To verify the effectiveness of MSCWMA in mine ventilation network optimization, an experimental mine was constructed, and its mine ventilation network topology diagram is shown below. Figure 10 As shown, the mine ventilation network consists of 19 nodes and 27 branches, including one intake shaft and one return shaft. Branch 26 is the branch where the ventilation fan is located.

[0102] The ventilation network of the experimental mine was calculated using MATLAB. The parameters of the roadway ventilation system before optimization and the calculated air volume are shown in Table 3 below.

[0103] Table 3 Parameters of the roadway ventilation system before optimization in the experimental mine

[0104] The air pressure characteristic curve of the ventilator is given by equation (25):

[0105] In the formula: — Fan air volume.

[0106] (ii) Adjustable branch set solution based on air volume sensitivity In the experimental mine, taking branch 8 as an example, when the real-time airflow of this branch fails to meet the ventilation network requirements due to excessive methane concentration, a ventilation adjustment scheme based on ventilation network sensitivity needs to be activated. This increases the air resistance of some branches, thereby increasing the airflow of that branch, thus reducing the methane concentration and ensuring that branch 8 is in a safe operating state. Table 3 shows that the initial airflow of branch 8 is 5.56 m³ / s. 3 / s, assuming that the gas sensor of branch 8 detects that the gas concentration of the branch increases to 1%, the absolute gas emission of the branch is calculated. The formula for calculating the absolute gas emission is as follows (26):

[0107] In the formula: m is the absolute gas emission rate. 3 / s; This refers to the real-time air volume in the tunnel. The methane concentration is expressed as %.

[0108] From equation (26), it can be seen that the absolute gas emission rate of branch 8 is 0.0556 m³ / s. 3 / s, based on the formula for the air volume required in the gas-emitting roadway, the minimum air volume required when the gas concentration exceeds the standard can be obtained, as shown in the following formula (27):

[0109] In the formula: The minimum air volume required for the roadway when the gas concentration exceeds the standard, in m 3 / s; For safety, a factor of 100 is usually used; The ventilation factor for operations typically ranges from [value range missing]. In this application, the experimental value is 1.5.

[0110] From equation (27), it can be seen that the minimum air volume required when the gas concentration in branch 8 exceeds the standard is 8.34 m³ / s. 3The maximum airflow rate is 1 / s. However, when the airflow is too high, the total amount of gas gushing out of the tunnel will also increase, potentially creating a safety hazard. Therefore, to ensure safety, the maximum airflow rate of branch 8 should be the minimum airflow rate required when the gas concentration exceeds the standard. This is 1.2 times the maximum adjustable airflow of branch 8, which is 10.01m³. 3 / s.

[0111] A program was written using MATLAB to solve the air volume sensitivity problem. The basic parameter data in Table 3 was input for the calculation. The sensitivity of all branches to branch 8 is shown in Table 4.

[0112] Table 4. Sensitivity of all branches to branch 8

[0113] The values ​​in Table 4 reflect the impact of wind resistance on air volume in each branch of the wind network. The degree of influence. This application adjusts the air volume of branch 8 by increasing resistance. Therefore, it is necessary to select branches with sensitivity greater than 0 for resistance adjustment. After sorting the sensitivity values ​​in descending order, the following branches are obtained: 16, 4, 9, 2, 24, 6, 14, 19, 13, 5, 11, 12. Since the sensitivity of branches 16, 4 and 9 is greater than 2, they have a greater influence on the air volume of branch 8. Therefore, branches 16, 4 and 9 are selected as the adjustable branch set.

[0114] To further determine the wind resistance adjustment range of each branch, it is necessary to analyze the airflow sensitivity attenuation characteristics of branches 16, 4 and 9. Ten sets of wind resistance values ​​were taken for each of the three branches and the corresponding sensitivity values ​​were calculated. The calculation results are shown in Table 5.

[0115] Table 5 Sensitivity d 8,j Wind resistance R j Change data

[0116] The data in Table 5 were fitted using a power function fitting analysis method. The sensitivity was obtained by processing the data using Matlab. and wind resistance The fitting curve of the change is as follows Figure 11 , Figure 12 and Figure 13 As shown.

[0117] Depend on Figure 11 The sensitivity can be calculated. and wind resistance The relationship is

[0118] Depend on Figure 12 The sensitivity can be calculated. and wind resistance The relationship is

[0119] Depend on Figure 13 The sensitivity can be calculated. and wind resistance The relationship is

[0120] Based on the analysis in Section 2 and combined with equation (8), the branches 16, 4, and 9 can be calculated respectively. Reasonable range for increasing resistance and adjusting airflow:

[0121]

[0122] .

[0123] (III) Analysis of Target Airflow Optimization Results Based on MSCWMA Algorithm Based on the nonlinear unconstrained objective function for ventilation network airflow regulation established in Section 1, this application will... The objective function to be solved is the adjustable branch set and its wind resistance adjustment range. After determining these, the MSCWMA is used to optimize the wind volume objective function.

[0124] To verify the superiority of MSCWMA in solving airflow optimization problems, five algorithms—ALA, WMA, WHO, WOA, and MSCWMA—were compared. To solve this problem, the number of branches was adjusted to 3 for each branch to determine... The maximum adjustment range, and the algorithm parameters are set as follows: population size Maximum number of iterations Variable dimensions To avoid randomness in the solution and reduce the randomness of the algorithm, all five algorithms were optimized 20 times. The optimization results are shown in Table 6.

[0125] Table 6. Effects of Algorithm Improvement Before and After Optimization results

[0126] As shown in Table 6, the MSCWMA affects the airflow of branch 8. The average optimized airflow and the optimal airflow solution obtained by MSCWMA are both superior to those of ALA, WMA, WHO, and WOA. In terms of runtime, MSCWMA has a longer convergence time, which is because the strategy employed enhances the algorithm's global search capability; the maximum optimized airflow in branch 8 can reach 10.61 m³ / s. 3 / s, at this time, the wind resistance of the branches is adjusted as follows: , , . Figure 14 The figure shows the iterative graph of the five algorithms for optimizing the air volume in branch 8. As can be seen from the figure, MSCWMA is superior to the other four algorithms in terms of global search capability and optimization effect, showing better exploration capability and effectively avoiding local optima. It has good superiority in solving practical problems.

[0127] According to the calculations by the air network, the air volumes of other branches after air adjustment are shown in Table 7. The data in the table shows that the air volumes of each branch after adjustment meet the minimum required air volume for that branch, thus complying with the air adjustment requirements. The adjustable air volume range for branch 8 is... m 3 / s, the air volume can be adjusted up to 90.83%, which can effectively solve the disaster problem when the gas concentration exceeds the limit in branch 8.

[0128] Table 7. Optimized airflow distribution for each branch

[0129] (iv) Determination of the optimal air conditioning scheme In actual operation, different numbers of regulating branches are selected for air adjustment based on the specific air volume required by the air-consuming branch. For branch 8, branch 16, which generally has the highest sensitivity value, is preferred. The airflow is adjusted within a certain range. If the target airflow is not met, the next optimal adjustment branch is selected sequentially to optimize the airflow until the requirement is met. Table 8 shows the effect of MSCWMA on different numbers of adjustment branches. Three optimization schemes.

[0130] Table 8. Effects of MSCWMA on different numbers of regulatory branches Optimization solution

[0131] As shown in Table 8, when the gas concentration in branch 8 increases to 1%, if only the resistance of branch 16 is adjusted, the maximum adjustable air volume of branch 8 is 7.61 m³. 3 The current airflow rate of / s is insufficient to meet the air demand of branch 8, so a multi-branch joint commissioning method must be adopted. When branches 16 and 4 are jointly commissioned and their resistance increased, the maximum airflow can be adjusted to 9.44m³ / s. 3 / s, higher than the minimum airflow of 8.34m³ required when the gas concentration in branch 8 exceeds the standard. 3 / s, while being lower than the maximum allowable airflow of 10.01m³ / s for branch 8. 3 / s, which meets the airflow requirements of branch 8. If the resistances of branches 16, 4, and 9 are adjusted simultaneously, the maximum adjustable airflow of branch 8 is 10.61m³. 3 / s, which is higher than the maximum allowable airflow of 10.01m³ / s for branch 8. 3 / s. Therefore, the optimal adjustment branch set for branch 8 is The wind resistance of the branches is adjusted as follows: , .

[0132] In conclusion, when , At that time, the air volume of branch 8 can reach 9.44m³. 3 / s, thus determining the optimal solution for this resistance-increasing airflow adjustment.

[0133] To improve the optimization capabilities of Whale Leader Search (WMA), this application proposes an MSCWMA strategy, which includes introducing a best-point set population initialization strategy, a golden sine algorithm update strategy, an SMA strategy, and a bidirectional population evolution dynamics strategy. The best-point set population initialization strategy makes the initial population distribution more uniform, improving population diversity and the algorithm's global search capability during the initialization phase. The golden sine algorithm update strategy in the whale leader update phase accelerates the algorithm's optimization speed, expanding the search range and enhancing its search capability. The SMA strategy in the whale follower update phase accelerates convergence, enhancing the algorithm's global optimization capability. The bidirectional population evolution dynamics strategy, applied to both the current leader and follower individuals, helps the algorithm escape local optima, thereby improving population diversity and search capability.

[0134] A smart decision-making model for on-demand control of mine ventilation volume was constructed, with the air volume of the demand branches in the mine ventilation network as the objective function. The optimal adjustable branch set and air resistance adjustment range were determined using air volume sensitivity theory. The models were solved using ALA, WMA, WHO, WOA, and MSCWMA to optimize the control of the demand branches. Experimental results show that the average optimized air volume using MSCWMA is 10.53 m³ / s. 3 / s, the optimal air volume is 10.61m³ / s. 3 / s, the airflow optimization results are superior to the other four algorithms. Therefore, MSCWMA has better optimization performance in terms of global search capability and optimization effect.

[0135] The implementation principle of this application is as follows: Based on mine ventilation theory, a ventilation network solution method is studied. With maximizing the adjustable airflow of the required branch as the optimization objective, a mine ventilation network airflow optimization model is established. A non-differentiable precise penalty function is used to transform the constraint problem in the optimization process, and the airflow adjustment optimization scheme is determined through airflow sensitivity theory. Addressing the shortcomings of WMA (Wide Motion Optimization) in solving complex optimization problems, this application improves WMA by proposing an MSCWMA algorithm, which is applied to the airflow optimization and adjustment of the ventilation network. ALA (Asynchronous Algorithm), WMA, WHO (Wide Motion Optimization), and WOA (Wide Motion Optimization) are compared. Experimental results show that MSCWMA outperforms the other four algorithms in both global search capability and optimization effect. When the gas concentration in the required branch exceeds the limit, after optimizing the ventilation network airflow adjustment model using MSCWMA, the airflow value of the required branch can be increased by 90.83% while ensuring a reasonable allocation of the minimum required airflow to other branches. This effectively solves the catastrophic problem when the gas concentration in the required branch exceeds the limit, demonstrating the superior optimization performance of the multi-strategy collaboratively improved whale migration algorithm.

[0136] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A method for optimizing and adjusting ventilation volume in mine roadways, characterized in that, include: S1: First, identify the target air demand branch that needs airflow optimization and adjustment. Calculate the airflow sensitivity matrix corresponding to the ventilation network to which the target air demand branch belongs using the airflow sensitivity calculation formula. In the airflow sensitivity matrix, sort the airflow sensitivity of each branch acting on the target air demand branch according to their numerical values. Use the resistance-increasing airflow adjustment method to adjust the airflow, selecting branches with airflow sensitivity greater than 0 as the adjustable branch set. Use regression analysis to establish a functional relationship between airflow sensitivity and roadway resistance, calculating the reasonable control range of the resistance parameter. Use a multi-strategy collaborative improvement whale migration optimization algorithm mathematical model to optimize the resistance value of the adjustable branches. S2: with To define the objective function for regulating airflow in the ventilation network, initialize population information and relevant parameters, including population size. Maximum number of iterations The variable dimension corresponding to the number of adjustable branches. The upper limit of the solution space corresponding to the adjustable branch wind resistance adjustment range and lower limit The adjustable branch's wind resistance adjustment value is The upper limit of the solution space corresponding to the adjustable branch wind resistance adjustment range and lower limit This refers to the reasonable adjustment range for wind resistance; S3: Initialize the population using the optimal point set method and randomly generate the initial population. ; S4: Substitution Calculate each individual air volume adaptability value Obtain the current maximum airflow fitness value and the location of the corresponding individual. And record it, among which The optimal solution for the air conditioning parameters represents the best air resistance adjustment value for the adjustable branch; S5: For each individual air volume adaptability value Sort the individuals and select the top 20% with the best fitness values ​​as the leaders of the population, and the bottom 80% with the worst fitness values ​​as the followers of the population. S6: After updating the current position of the population leader according to the leader's position update formula, the individual position is updated again using the golden sine algorithm strategy; the current position of the population followers is updated according to the improved follower position update formula. S7: After completing the individual position update, a two-way population evolution dynamics strategy is adopted to update the position of each whale individual; S8: Compare the airflow adaptability values ​​for each individual. Preserve the optimal airflow adaptation value and the corresponding optimal location. Determine if the current number of iterations reached is the maximum. If it is, execute S9; otherwise, return to execute S5. S9: Output the global optimal solution and the corresponding optimal airflow fitness value This yields the set of adjustable branches that can meet the airflow requirements of the branch requiring airflow, and the optimal air resistance adjustment of the adjustable branches. The optimal air resistance adjustment of the adjustable branches is the global optimal solution. This ensures that the airflow of the branch ventilation system meets the ventilation requirements.

2. The method for optimizing and adjusting ventilation volume in mine roadways according to claim 1, characterized in that: When optimizing the wind resistance value of adjustable branches using a multi-strategy collaborative improvement algorithm mathematical model for whale migration, the wind resistance value of the branch with the highest airflow sensitivity value in the adjustable branch set is optimized first. If the highest branch meets the airflow requirement of the branch requiring airflow, then the highest branch is the optimal adjustable branch set. If it does not meet the airflow requirement of the branch requiring airflow, then the wind resistance value of the branch with the second highest airflow sensitivity value in the adjustable branch set is adjusted. If it still does not meet the airflow requirement of the branch requiring airflow, then the wind resistance value of the branch with the third highest airflow sensitivity value in the adjustable branch set is adjusted. This process continues until the airflow requirement of the branch requiring airflow can be met, at which point these branches are the optimal adjustable branch set.

3. The method for optimizing and adjusting ventilation volume in mine roadways according to claim 1, characterized in that: In S3, the formulas for initializing the population individuals using the optimal point set method are shown in equations (14) and (15): The principle of optimal point set is based on the assumption that... The unit cube in Euclidean space There exists a set of points in the set, represented by equation (14): ； Its deviation satisfies ,in, It is only with The relevant constants, at this time, It's a collection of great spots. For the optimal point, its value is ,in, It is to satisfy The smallest prime number, mapped to the initial population, can be expressed as equation (15): 。 4. The method for optimizing and adjusting ventilation volume in mine roadways according to claim 1, characterized in that: In S6, after the population leader updates its current position according to the leader's position update formula, the golden sine algorithm strategy is used to update the individual positions again; the population followers update their current positions according to the improved follower position update formula. The formula for updating the position of the population leader is Equation (11): ； In the formula: This represents the current iteration number; For the first The whale in the The position at the next iteration; For the first The whale in the The position at the next iteration; and All Random numbers within the interval; For the number of leader whales, ; The formula (16) for updating the individual position again using the golden sine algorithm strategy is: ； In the formula: for The random number within the interval determines the distance an individual moves during the iteration; for Random numbers within the interval determine the direction of position updates for an individual during iteration; the golden ratio. , The golden ratio It is an irrational number; The improved follower position update formula (17) is as follows: ； In the formula: For the first The whale in the Fitness value at the next iteration; For the first The whale in the Fitness value at the next iteration; For interval The random number, where As shown in equation (18); 1 represents the adaptive weight, as shown in equation (19); and For individuals randomly selected from followers Two other different whales and whales Location; ； ； In the formula: This represents the maximum number of iterations.

5. The method for optimizing and adjusting ventilation volume in mine roadways according to claim 4, characterized in that: In S7, when the bidirectional population evolution dynamics strategy updates the position of each whale individual, new mutant individuals are obtained for each whale individual of the whale leader through natural mutation, as shown in Equation (20). ； The whale followers are randomly divided into two groups. One group of whales will mutate around the best individual in the group, as shown in Equation (21). The other group of whales will relocate to their initial position for further exploration, as shown in Equation (22). ； ； In the formula: For the first The location of an individual whale after mutation; The mutation factor is shown in equation (23); and For individuals randomly selected from the leaders The other two different whales and whales Location; , ; ； For each individual whale, retain the better individuals before and after the mutation, as shown in Equation (24): ； In the formula: For the first The fitness value of each individual whale after mutation; the position of the population leader is updated according to equations (20) and (24), and the random number is determined. If the value is less than 0.5, the follower position is updated according to equations (21) and (24); otherwise, the follower position is updated according to equations (22) and (24).

Images