Ventilation quantity intelligent algorithm based on air quality of excavation face of extra-long tunnel

Through artificial intelligence algorithms to calculate and adjust the ventilation volume of the excavation surface of the extra-length tunnel, the problem of difficulty in accurately determining the ventilation volume in the existing technology is solved, and efficient ventilation is achieved, reducing costs and improving construction safety.

CN118934056APending Publication Date: 2024-11-12CHINA TIESIJU CIVIL ENGINEERING GROUP CO LTD +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411153407.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-21
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

The prior art cannot accurately determine the ventilation volume required for the excavation surface of the extra-length tunnel, resulting in excessive fan power increasing costs, and it is difficult to discharge dirty air in small ventilation volume, affecting construction efficiency and safety.

Method used

Using the ventilation intelligent algorithm based on artificial intelligence, by obtaining ventilation parameters and dangerous gas concentration sets, calculate the return air wind speed, number of people and ventilation required for internal combustion engine, input it into the air volume prediction model, adjust the ventilation volume, consider the return air volume, number of people and Dangerous gas concentration and adjust ventilation volume in real time.

Benefits of technology

It has achieved the determination of ventilation volume based on actual needs, reducing fan power, improving ventilation efficiency, ensuring air quality, reducing costs, and improving construction safety.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118934056A_ABST
    Figure CN118934056A_ABST
Patent Text Reader

Abstract

The invention discloses an intelligent ventilation quantity algorithm based on air quality of an excavation face of an extra-long tunnel. The intelligent ventilation quantity algorithm comprises a data acquisition module, an air quantity calculation module and an air quantity adjustment module, relates to the technical field of tunnel ventilation quantity calculation, and solves the technical problem that the specific required ventilation quantity of a tunnel face cannot be determined according to actual requirements in an extra-long tunnel. According to the ventilation parameters, the ventilation quantity needed by the return air speed, the ventilation quantity needed by the number of people and the ventilation quantity needed by the internal combustion engine are obtained and input into a ventilation quantity prediction model together with the hazardous gas concentration set, and the needed ventilation quantity is obtained; in this way, the required ventilation quantity of the tunnel face is considered from the dimensions of the return air quantity in the tunnel, the number of workers and the concentration of turbid and hazardous gas generated by working of an internal combustion engine, and therefore the contaminated air of the tunnel face of the extra-long tunnel can be smoothly exhausted out of the tunnel.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of tunnel ventilation volume calculation, and specifically is an intelligent ventilation volume algorithm based on the air quality of the excavation surface of a super-long tunnel. Background Art

[0002] The construction and operation of extra-long tunnels (generally tunnels with a length of more than 10 kilometers) place extremely high demands on ventilation systems. During tunnel excavation, especially when using the drill and blast method or tunnel boring machine, a large amount of dust, hazardous gases (such as carbon monoxide, sulfur dioxide, etc.) and heat are generated, which not only affect the health and safety of construction workers, but also reduce visibility, affecting construction efficiency and equipment performance.

[0003] The current tunnel ventilation scheme is basically selected according to the air supply requirements during the most difficult ventilation period. The fans generally used are of large power, which increases the investment cost of the fans. Secondly, the ventilation volume generated by the large power is easy to raise dust and affect work efficiency when the drilling and blasting method is used for tunnel construction. It is difficult to ensure that the dirty air in the extra-long tunnel is discharged outside the tunnel if the ventilation volume is controlled at a small level. Therefore, it is currently impossible to accurately determine the specific ventilation volume required for the face. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an intelligent ventilation algorithm based on the air quality of the excavation face of an extra-long tunnel, which is used to solve the technical problem that the specific required ventilation volume of the face cannot be determined according to actual needs in the extra-long tunnel. The present invention obtains ventilation parameters, correction parameters and a set of hazardous gas concentrations, and obtains the ventilation volume required for the return air speed, the ventilation volume required for the number of people and the ventilation volume required for the internal combustion engine according to the ventilation parameters, and inputs them together with the hazardous gas concentration set into the ventilation volume prediction model to obtain the required ventilation volume. In this way, the required ventilation volume of the face is considered from the dimensions of the return air volume in the tunnel, the number of workers, the turbidity generated by the internal combustion engine and the hazardous gas concentration to solve the above-mentioned problem.

[0005] To achieve the above-mentioned object, a first aspect of the present invention provides an intelligent ventilation algorithm based on the air quality of the excavation face of a super-long tunnel, comprising: a data acquisition module and an air volume calculation module;

[0006] The data acquisition module is used to obtain ventilation parameters, correction parameters and dangerous gas concentration set of the current node during tunnel excavation; wherein the ventilation parameters include: tunnel excavation section, maximum number of workers at the tunnel face, internal combustion engine power and minimum wind speed at the tunnel face; the correction parameters include inclined shaft length, main tunnel excavation length and air duct leakage rate;

[0007] The air volume calculation module: based on the minimum wind speed of the tunnel face in the ventilation parameters and the tunnel excavation section, obtains the ventilation volume required for the return air speed; based on the maximum number of workers on the tunnel face in the ventilation parameters, obtains the ventilation volume required for the number of people; based on the internal combustion engine power in the ventilation parameters, obtains the ventilation volume required for the internal combustion engine; and,

[0008] The required ventilation volume for the return air speed, the required ventilation volume for the number of people, the required ventilation volume for the internal combustion engine and the concentration of hazardous gases are input into the air volume prediction model to obtain the reference ventilation volume; wherein the air volume prediction model is obtained based on an artificial intelligence model;

[0009] The air volume adjustment module: adjusts the reference air volume based on the correction parameter to obtain the required ventilation volume; and adjusts the required ventilation volume in real time according to the concentration of dangerous gases at the current node;

[0010] It should be noted that the power of the internal combustion engine is the operating power of equipment such as excavators, loaders, and slag trucks working in the factory.

[0011] Preferably, the calculation formula for the ventilation volume required for the return air speed is: QX1=V×S; wherein V is the minimum wind speed at the face, S is the tunnel excavation section, and QX1 is the ventilation volume required for the return air speed.

[0012] Preferably, the calculation formula for the ventilation volume required for the number of people is: QX2=α×K×N; wherein α is the air supply standard per person, 10>α>3, K is the tunnel ventilation coefficient, N is the maximum number of workers on the face, and QX2 is the ventilation volume required for the number of people.

[0013] It should be noted that the tunnel ventilation coefficient K is determined based on factors such as the tunnel's air leakage and uneven distribution, and the value range of K is [0.8, 1.2].

[0014] Preferably, the required ventilation volume QX3 of the internal combustion engine is: Where W i is the power of the i-th internal combustion engine, η i is the effective utilization rate of the i-th internal combustion engine, v is the required wind speed corresponding to the unit power of the internal combustion engine, and n is the total number of internal combustion engines.

[0015] It should be noted that η i The value range is [0.2, 0.8], and the required wind speed v corresponding to the unit power of the internal combustion engine is a constant, generally v = 0.05m 3 / s.

[0016] Preferably, the air volume prediction model is obtained based on an artificial intelligence model, including:

[0017] Extract the required ventilation volume for return air speed, the required ventilation volume for the number of people, the required ventilation volume for the internal combustion engine, the dangerous gas concentration at the face and the reference ventilation volume from the historical database; and construct the training set and the test set for model training and testing; the ratio of the training set to the test set is 8:2;

[0018] Based on the recurrent neural network and introducing the attention mechanism, the wind volume prediction model to be trained is constructed, the wind volume prediction model to be trained is trained through the training set, and then the completed training wind volume prediction model is tested through the test set;

[0019] The accuracy is verified according to the model accuracy calculation formula: PR=TP / (TP+FR), and the trained air volume prediction model is adjusted according to the accuracy verification result. Finally, the air volume prediction model with the input of the required ventilation volume of return air speed, the required ventilation volume of the number of people, the required ventilation volume of the internal combustion engine, and the concentration of dangerous gases on the face is obtained, and the output is the benchmark ventilation volume. Among them, PR is the accuracy value, TP is the amount of data detected correctly, and TP is the amount of data detected incorrectly.

[0020] Preferably, the adjusting the reference air volume based on the correction parameter includes:

[0021] Extract the correction parameters and reference ventilation volume, and use the formula QX0 = Qz / (1-β) (L1+L2) / 1000 Calculate the required ventilation volume QX0; where Qz is the reference ventilation volume, β is the air duct leakage rate, L1 is the length of the inclined shaft, and L2 is the main tunnel excavation length.

[0022] The above formula can take into account the situation where air leakage occurs during the air delivery process, resulting in a reduction in ventilation volume compared to actual demand.

[0023] Preferably, the real-time adjustment of the required ventilation volume according to the dangerous gas concentration at the current node includes:

[0024] S1: extracting the hazardous gas concentration set of the current node and the hazardous gas concentration set of the previous node, and obtaining the concentration change coefficient based on the hazardous gas concentration set of the previous node and the hazardous gas concentration set of the current node;

[0025] S2: Determine whether the concentration change coefficient is greater than the change threshold; if yes, jump to S3; if no, maintain the current required ventilation volume;

[0026] S3: Matching the concentration change coefficient with the concentration change-air volume level mapping table to obtain the increased air volume;

[0027] S4: Calculate the sum of the required ventilation volume and the increased air volume to obtain the adjusted required ventilation volume.

[0028] It should be noted that the concentration change-air volume level mapping table is obtained based on the experience of those skilled in the art, and the larger the concentration change coefficient, the larger the corresponding increased air volume.

[0029] Preferably, the step of obtaining the concentration variation coefficient based on the hazardous gas concentration set of the previous node and the hazardous gas concentration set of the current node includes:

[0030] The concentration values ​​of each hazardous gas in the hazardous gas concentration set of the previous node are extracted and marked as NQj, and the concentration values ​​of each hazardous gas in the hazardous gas concentration set of the current node are extracted and marked as NHj; wherein NQj represents the concentration value of the j-th hazardous gas in the previous node, and NHj represents the concentration value of the j-th hazardous gas in the current node;

[0031] pass Calculate the concentration variation coefficient, where m is the total amount of hazardous gas and μ j is the weight coefficient of the jth hazardous gas, and

[0032] Compared with the prior art, the present invention has the following beneficial effects:

[0033] 1. In the present invention, by obtaining ventilation parameters, correction parameters and hazardous gas concentration sets, the ventilation volume required for the return air speed, the ventilation volume required for the number of people and the ventilation volume required for the internal combustion engine are respectively obtained according to the ventilation parameters, and they are input into the ventilation volume prediction model together with the hazardous gas concentration set to obtain the required ventilation volume. In this way, the required ventilation volume of the face is considered from the dimensions of the return air volume in the tunnel, the number of workers, the turbidity generated by the internal combustion engine and the hazardous gas concentration. In this way, the dirty air of the face of the extra-long tunnel can be smoothly discharged out of the tunnel, and a fan of suitable power can be selected according to the required ventilation volume and the speed of the fan can be adjusted to reduce the cost.

[0034] 2. In the present invention, when obtaining the hazardous gas concentration set, the hazardous gas concentration sets of multiple nodes are obtained in a continuous manner. When the hazardous gas concentration set that is most likely to occur in the face of the extra-long tunnel changes, the current required ventilation volume is adjusted according to the change in the concentration of each gas, thereby improving the work safety of the staff in the extra-long tunnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0036] Figure 1It is a flow chart of the intelligent algorithm of ventilation volume based on the air quality of the excavation face of a super-long tunnel according to the present invention;

[0037] Figure 2 This is a logic judgment block diagram for obtaining the required ventilation volume in the present invention;

[0038] Figure 3 The present invention is a logic block diagram of the ventilation volume intelligent algorithm based on the air quality of the excavation surface of a super-long tunnel. DETAILED DESCRIPTION

[0039] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0040] See also Figure 1 and Figure 3 As shown, the first embodiment of the present invention provides a ventilation volume intelligent algorithm based on the air quality of the excavation face of a super-long tunnel, including: a data acquisition module and an air volume calculation module;

[0041] The data acquisition module is used to obtain ventilation parameters, correction parameters and dangerous gas concentration set of the current node during tunnel excavation; wherein the ventilation parameters include: tunnel excavation section, maximum number of workers at the tunnel face, internal combustion engine power and minimum wind speed at the tunnel face; the correction parameters include inclined shaft length, main tunnel excavation length and air duct leakage rate;

[0042] The air volume calculation module: based on the minimum wind speed of the tunnel face in the ventilation parameters and the tunnel excavation section, obtains the ventilation volume required for the return air speed; based on the maximum number of workers on the tunnel face in the ventilation parameters, obtains the ventilation volume required for the number of people; based on the internal combustion engine power in the ventilation parameters, obtains the ventilation volume required for the internal combustion engine; and,

[0043] The required ventilation volume for the return air speed, the required ventilation volume for the number of people, the required ventilation volume for the internal combustion engine and the concentration of hazardous gases are input into the air volume prediction model to obtain the reference ventilation volume; wherein the air volume prediction model is obtained based on an artificial intelligence model;

[0044] The air volume adjustment module: adjusts the reference air volume based on the correction parameter to obtain the required ventilation volume; and adjusts the required ventilation volume in real time according to the concentration of dangerous gases at the current node;

[0045] In this embodiment, the air volume prediction model is obtained based on an artificial intelligence model, including:

[0046] Extract the required ventilation volume for return air speed, the required ventilation volume for the number of people, the required ventilation volume for the internal combustion engine, the dangerous gas concentration at the face and the reference ventilation volume from the historical database; and construct the training set and the test set for model training and testing; the ratio of the training set to the test set is 8:2;

[0047] Based on the recurrent neural network and introducing the attention mechanism, the wind volume prediction model to be trained is constructed, the wind volume prediction model to be trained is trained through the training set, and then the completed training wind volume prediction model is tested through the test set;

[0048] The accuracy is verified according to the model accuracy calculation formula: PR = TP / (TP+FR). When PR ≤ 95%, the trained air volume prediction model is adjusted and retrained. When PR ≥ 95%, the input is the ventilation volume required for the return air speed, the ventilation volume required for the number of people, the ventilation volume required for the internal combustion engine, and the concentration of dangerous gases on the face. The output is the air volume prediction model with the benchmark ventilation volume. Among them, PR is the accuracy value, TP is the amount of data detected correctly, and TP is the amount of data detected incorrectly.

[0049] For example, the following ventilation parameters exist: the minimum wind speed at the face is 0.25 m / s, and the tunnel excavation section is 71.8 m 2 The maximum number of workers on the face is 60. The internal combustion engine power of one excavator in the tunnel is 118kw, the internal combustion engine power of one loader is 162kw, the internal combustion engine power of two slag trucks is 140kw, and the internal combustion engine power of two concrete tankers is 250kw.

[0050] Corrected parameters: inclined shaft length L1 is 1330m, main tunnel excavation length L2 is 1950m, and air duct leakage rate is 1%;

[0051] The calculation formula for the ventilation volume required for the corresponding return air speed is: QX1 = V × S = 71.8 × 0.25 = 17.95 m 3 / s.

[0052] The formula for calculating the ventilation volume required for the number of people is: QX2 = α × K × N = 4 × 1.2 × 60 / 60 = 4.8m 3 / s; where α is the air supply standard per person per portion, α=4m 3 / min, K is the tunnel ventilation coefficient, and K = 1.2.

[0053] Among them, the tunnel ventilation coefficient K is determined based on factors such as the tunnel's air leakage and uneven distribution. When converting units, in order to convert minutes into seconds, it needs to be divided by 60 in the actual processing process.

[0054] The required ventilation volume of the internal combustion engine QX3: Where Wi is the power of the i-th internal combustion engine, η i is the effective utilization rate of the i-th internal combustion engine, v is the required wind speed corresponding to the unit power of the internal combustion engine, and n is the total number of internal combustion engines. For example, when the effective utilization rate of all internal combustion engines is 0.55, v = 0.05m 3 / s, then QX3=(118+162+2×140+2×250)×0.05×0.55=29.2m 3 / s.

[0055] The required ventilation volume of the return air speed, the required ventilation volume of the number of people, the required ventilation volume of the internal combustion engine and the concentration of dangerous gases are input into the air volume prediction model to obtain the reference ventilation volume of 32.7m 3 / s;

[0056] The reference air volume is adjusted based on the correction parameters, including:

[0057] Extract the correction parameters and reference ventilation volume, and use the formula QX0 = Qz / (1-β) (L1+L2) / 100 =32.7 / (1-0.01) (1330+1950) / 1000 =45.5m 3 / sCalculate the required ventilation volume QX0.

[0058] See also Figure 2 As shown, the required ventilation volume is adjusted in real time according to the dangerous gas concentration at the current node, including:

[0059] S1: extracting the hazardous gas concentration set of the current node and the hazardous gas concentration set of the previous node, and obtaining the concentration change coefficient based on the hazardous gas concentration set of the previous node and the hazardous gas concentration set of the current node;

[0060] S2: Determine whether the concentration change coefficient is greater than the change threshold; if yes, jump to S3; if no, maintain the current required ventilation volume;

[0061] S3: Matching the concentration change coefficient with the concentration change-air volume level mapping table to obtain the increased air volume;

[0062] S4: Calculate the sum of the required ventilation volume and the increased air volume to obtain the adjusted required ventilation volume.

[0063] It should be noted that the concentration change-air volume level mapping table is obtained based on the experience of those skilled in the art, and the larger the concentration change coefficient, the larger the corresponding increased air volume.

[0064] Specifically, the concentration variation coefficient is obtained based on the hazardous gas concentration set of the previous node and the hazardous gas concentration set of the current node, including:

[0065] The concentration values ​​of each hazardous gas in the hazardous gas concentration set of the previous node are extracted and marked as NQj, and the concentration values ​​of each hazardous gas in the hazardous gas concentration set of the current node are extracted and marked as NHj; wherein NQj represents the concentration value of the j-th hazardous gas in the previous node, and NHj represents the concentration value of the j-th hazardous gas in the current node;

[0066] pass Calculate the concentration variation coefficient, where m is the total amount of hazardous gas and μ j is the weight coefficient of the jth hazardous gas, and

[0067] The harmful gas concentration set in this embodiment includes: carbon monoxide, nitrogen oxides converted into nitrogen dioxide, sulfur dioxide, and hydrogen sulfide; and the specific weight coefficient of the jth dangerous gas can be formulated with reference to the allowable concentration in the tunnel as follows, and the concentration of carbon monoxide needs to be ≤30mg / m 3 , the concentration of nitrogen dioxide needs to be ≤5mg / m 3 , the concentration of sulfur dioxide needs to be ≤15mg / m 3 , the concentration of hydrogen sulfide needs to be ≤10mg / m 3 ;

[0068] In other embodiments, the concentration of harmful gases may also include methane, alcohols (acrolein), and carbon dioxide gases.

[0069] The concentration change-air volume level mapping table is shown in Table 1:

[0070]

[0071]

[0072] It should be noted that when the concentration variation coefficient is greater than 10, the fan needs to operate at maximum power and an alarm signal needs to be issued, and when the concentration variation coefficient is less than 1, the current ventilation volume can be maintained.

[0073] Part of the data in the above formula is calculated by removing the dimension and taking its numerical value. The formula is a formula closest to the actual situation obtained by software simulation of a large amount of collected data; the preset parameters and preset thresholds in the formula are set by technical personnel in this field according to actual conditions or obtained through simulation of a large amount of data.

[0074] The working principle of the present invention is as follows: the ventilation parameters, correction parameters and hazardous gas concentration set of the current node are obtained during tunnel excavation; the ventilation volume required for the return air speed, the ventilation volume required for the number of people and the ventilation volume required for the internal combustion engine are respectively obtained based on the ventilation parameters, and the ventilation volume required for the return air speed, the ventilation volume required for the number of people, the ventilation volume required for the internal combustion engine and the hazardous gas concentration set are input into the air volume prediction model to obtain the reference ventilation volume, and the reference air volume is adjusted based on the correction parameters to obtain the required ventilation volume; according to the hazardous gas concentration of the current node, it is judged whether to adjust the required ventilation volume in real time; if not, the required ventilation volume is maintained; if yes, the hazardous gas concentration set of the current node and the hazardous gas concentration set of the previous node are extracted, and the concentration change coefficient is obtained based on the hazardous gas concentration set of the previous node and the hazardous gas concentration set of the current node; it is judged whether the concentration change coefficient is greater than the change threshold; if not, the current required ventilation volume is maintained; if yes, the concentration change coefficient is matched with the concentration change-air volume level mapping table to obtain the increased air volume; and the sum of the required ventilation volume and the increased air volume is calculated to obtain the adjusted required ventilation volume.

[0075] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An intelligent ventilation algorithm based on the air quality of the excavation face of a very long tunnel, characterized in that: include: Data acquisition module, air volume calculation module and air volume adjustment module; The data acquisition module is used to obtain ventilation parameters, correction parameters and dangerous gas concentration set of the current node during tunnel excavation; wherein the ventilation parameters include: tunnel excavation section, maximum number of workers at the tunnel face, internal combustion engine power and minimum wind speed at the tunnel face; the correction parameters include inclined shaft length, main tunnel excavation length and air duct leakage rate; The air volume calculation module: based on the minimum wind speed of the tunnel face in the ventilation parameters and the tunnel excavation section, obtains the ventilation volume required for the return air speed; based on the maximum number of workers on the tunnel face in the ventilation parameters, obtains the ventilation volume required for the number of people; based on the internal combustion engine power in the ventilation parameters, obtains the ventilation volume required for the internal combustion engine; and, The required ventilation volume for the return air speed, the required ventilation volume for the number of people, the required ventilation volume for the internal combustion engine and the concentration of hazardous gases are input into the air volume prediction model to obtain the reference ventilation volume; wherein the air volume prediction model is obtained based on an artificial intelligence model; The air volume adjustment module: adjusts the reference air volume based on the correction parameter to obtain the required ventilation volume; and adjusts the required ventilation volume in real time according to the hazardous gas concentration of the current node.

2. According to claim 1, the ventilation volume intelligent algorithm based on the air quality of the excavation face of a super-long tunnel is characterized in that: The calculation formula for the ventilation volume required for the return air speed is: QX1=V×S; wherein V is the minimum wind speed at the tunnel face, S is the tunnel excavation section, and QX1 is the ventilation volume required for the return air speed.

3. The intelligent ventilation algorithm based on the air quality of the excavation face of a super-long tunnel according to claim 1 is characterized in that: The calculation formula for the ventilation volume required for the number of people is: QX2=α×K×N; where α is the air supply standard per person, 10>α>3, K is the tunnel ventilation coefficient, N is the maximum number of workers on the face, and QX2 is the ventilation volume required for the number of people.

4. The intelligent ventilation algorithm based on the air quality of the excavation face of a super-long tunnel according to claim 1 is characterized in that: The required ventilation volume QX3 of the internal combustion engine: Where W i is the power of the i-th internal combustion engine, η i is the effective utilization rate of the i-th internal combustion engine, v is the required wind speed corresponding to the unit power of the internal combustion engine, and n is the total number of internal combustion engines.

5. The intelligent ventilation algorithm based on the air quality of the excavation face of a super-long tunnel according to claim 1 is characterized in that: The wind volume prediction model is obtained based on an artificial intelligence model, including: Extract the required ventilation volume for return air speed, the required ventilation volume for the number of people, the required ventilation volume for the internal combustion engine, the concentration of dangerous gases at the face and the reference ventilation volume from the historical database; and construct the training set and test set for model training and testing; Based on the recurrent neural network and introducing the attention mechanism, the wind volume prediction model to be trained is constructed, the wind volume prediction model to be trained is trained through the training set, and then the completed training wind volume prediction model is tested through the test set; The accuracy is verified according to the model accuracy calculation formula: PR=TP / (TP+FR), and the trained air volume prediction model is adjusted according to the accuracy verification result. Finally, the air volume prediction model with the input of the required ventilation volume of return air speed, the required ventilation volume of the number of people, the required ventilation volume of the internal combustion engine, and the concentration of dangerous gases on the face is obtained, and the output is the benchmark ventilation volume. Among them, PR is the accuracy value, TP is the amount of data detected correctly, and TP is the amount of data detected incorrectly.

6. The intelligent ventilation algorithm based on the air quality of the excavation face of a super-long tunnel according to claim 1 is characterized in that: The adjusting the reference air volume based on the correction parameter includes: Extract the correction parameters and reference ventilation volume, and use the formula QX0 = Qz / (1-β) (L1+L2) / 100 Calculate the required ventilation volume QX0; where Qz is the reference ventilation volume, β is the air duct leakage rate, L1 is the length of the inclined shaft, and L2 is the main tunnel excavation length.

7. The intelligent ventilation algorithm based on the air quality of the excavation face of a super-long tunnel according to claim 1 is characterized in that: The real-time adjustment of the required ventilation volume according to the dangerous gas concentration at the current node includes: S1: extracting the hazardous gas concentration set of the current node and the hazardous gas concentration set of the previous node, and obtaining the concentration change coefficient based on the hazardous gas concentration set of the previous node and the hazardous gas concentration set of the current node; S2: Determine whether the concentration change coefficient is greater than the change threshold; if yes, jump to S3; if no, maintain the current required ventilation volume; S3: Matching the concentration change coefficient with the concentration change-air volume level mapping table to obtain the increased air volume; S4: Calculate the sum of the required ventilation volume and the increased air volume to obtain the adjusted required ventilation volume.

8. The intelligent ventilation algorithm based on the air quality of the excavation face of a super-long tunnel according to claim 7 is characterized in that: The step of obtaining the concentration variation coefficient based on the hazardous gas concentration set of the previous node and the hazardous gas concentration set of the current node includes: The concentration values ​​of each hazardous gas in the hazardous gas concentration set of the previous node are extracted and marked as NQj, and the concentration values ​​of each hazardous gas in the hazardous gas concentration set of the current node are extracted and marked as NHj; wherein NQj represents the concentration value of the j-th hazardous gas in the previous node, and NHj represents the concentration value of the j-th hazardous gas in the current node; pass Calculate the concentration variation coefficient, where m is the total amount of hazardous gas and μ j is the weight coefficient of the jth hazardous gas, and