Extra-long tunnel pumping and pressing combined mixed ventilation system and construction method

Abstract

The invention discloses a pumping and pressing combined mixed ventilation system for an extra-long tunnel and a construction method, and belongs to the technical field of tunnel ventilation. The extra-long tunnel comprises a flat pilot tunnel, a left tunnel and a right tunnel. The mixed ventilation system comprises a first axial flow fan, a second axial flow fan, a third axial flow fan, an air draft device, a detection device and a control device. Wherein the first axial flow fan is used for supplying fresh air to a tunnel face of a flat pilot tunnel, the second axial flow fan is used for supplying fresh air to a tunnel face of a left tunnel, the third axial flow fan is used for supplying fresh air to a tunnel face of a right tunnel, and the air draft device is used for exhausting dirty air in the left tunnel out of the tunnel; the detection device is used for detecting air parameters in the flat pilot tunnel, the left tunnel and the right tunnel, and the control device is used for controlling the first axial flow fan, the second axial flow fan, the third axial flow fan and the air draft device according to the detected air parameters. According to the method, a mixed ventilation mode combining pumping and pressing is adopted, the problem that the ventilation efficiency of an extra-long tunnel is greatly reduced along with deep tunneling is solved, and the cost is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of tunnel ventilation, and in particular relates to a combined extraction and pressure ventilation system and a construction method for an extra-long tunnel. Background Art

[0002] With the continuous advancement of global urbanization, the population concentration is increasing, and the demand for travel between cities is growing. As an important channel connecting cities, tunnel engineering plays a key role in transportation. At present, highway tunnels have entered a period of great development. According to the design standards of expressways, the construction of tunnels in mountainous and hilly areas and long tunnel groups has become an inevitable trend. When building roads in mountainous areas, tunnel design and construction are also increasing. The dirty and high temperature environment during tunnel construction will pose a serious threat to the health of construction workers and the normal operation of construction machinery. As the only means of air circulation inside and outside the tunnel, construction ventilation can cool the cave, provide fresh air, remove dust and toxic and harmful gases during tunnel construction, maintain the normal operation of construction machinery and equipment, and ensure the health and safety of construction workers. It is the lifeline of the safe construction of the entire tunnel project.

[0003] Traditional ventilation methods expose many drawbacks when the tunnel construction length exceeds a certain range. For example, when the tunnel construction exceeds 2000m, the air quality and air volume are significantly reduced due to comprehensive factors such as efficiency and air duct management; after exceeding 2500m, relay ventilation is generally required, and jet fans are added every 1000m to assist in exhaust, which greatly increases the ventilation cost and makes it difficult to effectively solve the problems of poor air quality at the face and high temperature during the pouring of the second lining. Therefore, exploring more efficient, energy-saving and ventilation technologies that can adapt to the needs of extra-long tunnel construction has become an important issue that needs to be solved in the field of tunnel engineering. Summary of the invention

[0004] In view of the above-mentioned shortcomings of the prior art, the present invention provides a hybrid ventilation system and construction method combining extraction and pressure for an extra-long tunnel.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0006] The first aspect of the present invention discloses a combined ventilation system of an extra-long tunnel with extraction and pressure, wherein the extra-long tunnel comprises a flat guide tunnel, a left tunnel and a right tunnel, an auxiliary construction passage is provided between the left tunnel and the flat guide tunnel, a vehicle passage and a pedestrian passage are provided between the right tunnel and the left tunnel, and the entrance of the left tunnel, the pedestrian passage, the auxiliary construction passages except the auxiliary construction passage farthest from the entrance, and the vehicle passages except the vehicle passage farthest from the entrance are closed;

[0007] The hybrid ventilation system includes a first axial flow fan installed in the parallel adit, a second axial flow fan and a third axial flow fan installed in the right adit, an exhaust device, a detection device and a control device installed in the left adit. Detection devices are installed in the parallel adit, the left adit and the right adit.

[0008] Among them, the first axial flow fan is used to supply fresh air to the face of the parallel adit through a ventilation duct, the second axial flow fan is used to supply fresh air to the face of the left adit through a ventilation duct, the third axial flow fan is used to supply fresh air to the face of the right adit through a ventilation duct, and the exhaust device is used to discharge the polluted air in the left adit out of the tunnel, so as to form a circulating air flow with the parallel adit and the right adit supplying air and the left adit exhausting air.

[0009] Among them, the detection device is used to detect the air parameters in the parallel adit, the left adit and the right adit, and the control device is used to control the first axial flow fan, the second axial flow fan, the third axial flow fan and the exhaust device according to the detected air parameters.

[0010] Furthermore, the first axial flow fan is located between the opening of the first construction auxiliary passage and the entrance of the parallel adit, the second axial flow fan is located between the opening of the first vehicle passage and the entrance of the right adit, and the third axial flow fan is located between the opening of the first vehicle passage and the entrance of the right adit. Among them, the first construction auxiliary passage is the construction auxiliary passage farthest from the entrance of the parallel adit, and the first vehicle passage is the vehicle passage farthest from the entrance of the right adit.

[0011] Furthermore, the exhaust device includes a first exhaust fan and a second exhaust fan, and the first exhaust fan and the second exhaust fan are backup fans for each other.

[0012] Furthermore, the detection device includes an air velocity sensor, a temperature sensor, a negative pressure sensor, a carbon monoxide sensor, a carbon dioxide sensor, an oxygen concentration sensor and a hydrogen sulfide sensor. The temperature sensor, the carbon monoxide sensor, the carbon dioxide sensor, the oxygen concentration sensor and the hydrogen sulfide sensor are installed at the face, the air velocity sensor is installed at the air outlet of the ventilation duct, and the negative pressure sensor is installed at the middle position between the face and the corresponding entrance.

[0013] Furthermore, the control device includes a data calculation unit, a fan control unit and a risk alarm unit. The data calculation unit is used to analyze the historical monitoring data and real-time monitoring data collected by the detection device by using a machine learning algorithm. The fan control unit is used to adjust the operating parameters of the first axial flow fan, the second axial flow fan, the third axial flow fan and the exhaust device according to the analysis result of the data calculation unit. The risk alarm unit is used to give a risk warning according to the detection result of the detection device.

[0014] Furthermore, analyzing the historical monitoring data and real-time monitoring data collected by the detection device by using a machine learning algorithm includes:

[0015] Perform data preprocessing on the real-time monitoring data to obtain a time series matrix; input the time series matrix into a pre-trained hybrid neural network model to obtain the predicted values of the air volume demand index Q d , the negative pressure demand index P d and the fan speed demand index N d ;

[0016] Among them, the data preprocessing method includes: removing the values in the monitoring data that exceed the preset standard range as outliers. The monitoring data includes wind speed v, temperature T, negative pressure P, carbon monoxide concentration C CO , carbon dioxide concentration oxygen concentration and hydrogen sulfide concentration Use the principal component analysis method to extract features and reduce the dimension of the monitoring data, and convert the original seven-dimensional data into a three-dimensional principal component feature vector X = [x1, x2, x3]; construct a sliding time window data sequence. According to the standard of one window every 10 minutes, 60 data points in each window, and a monitoring frequency of once per second, arrange the principal component feature vectors corresponding to these data points in sequence to form a time series matrix.

[0017] Furthermore, the hybrid neural network model includes a convolutional neural network, a long short-term memory network, and an output network;

[0018] The convolutional neural network includes two convolutional layers and a max pooling layer. The input of the convolutional neural network is the time series matrix; the convolutional kernel size of the first convolutional layer is 3×3, the stride is 1, the number of channels is 16, and the ReLU activation function is used; the convolutional kernel size of the second convolutional layer is 3×3, the stride is 1, the number of channels is 32, and the ReLU activation function is used; the pooling window size of the max pooling layer is 2×2;

[0019] The long short-term memory network includes two layers, and each layer has 64 hidden units. The input of the long short-term memory network is the feature sequence output by the convolutional neural network;

[0020] The output network includes an output layer and a fully connected layer before the output layer; the output layer has three nodes, and the three nodes respectively correspond to the air volume demand index Q for fan control d , the negative pressure demand index P d and the fan speed demand index N d ; the number of nodes in the fully connected layer is 128, and the ReLU activation function is used.

[0021] Furthermore, the training method of the hybrid neural network model includes:

[0022] Perform data preprocessing on the historical monitoring data to obtain a time series matrix; divide the time series matrix into a training set, a validation set, and a test set;

[0023] Set the total number of training cycles. In each training cycle, take out the time series matrix sample data from the training set one by one, input it into the hybrid neural network model, and obtain the predicted output value y i , corresponding to the air volume demand index Q for fan control d , negative pressure demand index P d and fan speed demand index N d .

[0024] Furthermore, adjust the operating parameters of the first axial flow fan, the second axial flow fan, the third axial flow fan and the exhaust device according to the analysis results of the data calculation unit, including:

[0025] Determine the collaborative control coefficient α according to the change rate ΔC of the carbon monoxide concentration detected by the detection device corresponding to the axial flow fan CO and air volume demand index Q d ; g

[0026] Calculate the target speed N of the first axial flow fan, the second axial flow fan and the third axial flow fan target , and the calculation formula is:

[0027]

[0028] where k1 is the speed-air volume ratio coefficient of the axial flow fan;

[0029] Calculate the air density ρ according to the temperature T and negative pressure P in the tunnel, and the calculation formula is:

[0030]

[0031] where P0 is the standard physical atmospheric pressure and T0 is the absolute temperature under standard conditions;

[0032] The calculation formula of the air volume Q is:

[0033]

[0034] where k2 is a fixed coefficient and f(θ) is a function of the blade angle;

[0035] According to the air volume demand index Q d , negative pressure demand index P d , air density ρ and the performance curve of the exhaust fan, use the particle swarm optimization algorithm to obtain the optimal speed n target and blade angle θ target .

[0036] The second aspect of the present invention discloses a construction method, including: ​

[0037] S100. When the driving distance is less than the first preset value, forced ventilation is carried out. The method of forced ventilation is as follows:

[0038] Set a first axial flow fan at the entrance of the pilot tunnel, and the first axial flow fan sends fresh air to the face of the pilot tunnel through an air duct.

[0039] Set a second axial flow fan at the entrance of the left tunnel, and the second axial flow fan sends fresh air to the face of the left tunnel through an air duct.

[0040] Set a third axial flow fan at the entrance of the right tunnel, and the third axial flow fan sends fresh air to the face of the right tunnel through an air duct.

[0041] S200. When the driving distance is greater than or equal to the first preset value, hybrid ventilation is carried out. The method of hybrid ventilation is as follows:

[0042] Set the first axial flow fan, the second axial flow fan, the third axial flow fan, the exhaust device, the detection device and the control device according to the hybrid ventilation system described in the first aspect of the present invention.

[0043] Seal the entrance of the left tunnel, all pedestrian passages, all auxiliary construction passages except the auxiliary construction passage farthest from the entrance, and all vehicle passages except the vehicle passage farthest from the entrance.

[0044] Wherein, every time the length of the second preset value is advanced forward, the first axial flow fan, the second axial flow fan, the third axial flow fan and the detection device are moved forward a preset distance along the driving direction.

[0045] The technical solution provided by the present invention has the following beneficial effects compared with the known prior art:

[0046] (1) The present invention adopts a hybrid ventilation method combining exhaust and pressure, uses the already penetrated tunnel as an air duct to form a negative pressure channel, generates a negative pressure difference, and forms a circulating air flow, effectively solving the problem of air supply for ultra-long distance single-heading driving, overcoming the problem that the ventilation efficiency of extra-long tunnels decreases significantly as the driving progresses, ensuring the ventilation quality, helping to maintain the normal progress of construction, and reducing energy consumption.

[0047] (2) Through the innovative neural network structure, axial flow fan group collaborative control and exhaust fan adaptive optimization algorithm, the present invention can adjust the fan operation parameters more intelligently and accurately according to the tunnel ventilation requirements, improve the efficiency and stability of the ventilation system, while reducing energy consumption and ensuring construction safety. Description of the Drawings

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0049] Figure 1 It is a schematic diagram of a hybrid ventilation system in the present invention;

[0050] Figure 2 It is a schematic diagram of forced ventilation;

[0051] In the figure, 1 - horizontal pilot tunnel, 2 - left tunnel, 3 - right tunnel, 4 - auxiliary construction passage, 5 - vehicle passage, 6 - first axial flow fan, 7 - second axial flow fan, 8 - third axial flow fan, 9 - exhaust device, 10 - air duct, 11 - tunnel face. Detailed implementation manners

[0052] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0053] The following further describes the present invention with reference to embodiments.

[0054] As Figures 1 to 2 shown, this embodiment discloses a hybrid ventilation system combining exhaust and pressure in a super-long tunnel.

[0055] The first aspect of this embodiment discloses a hybrid ventilation system combining exhaust and pressure in a super-long tunnel. As Figure 1 shown, the super-long tunnel includes a horizontal pilot tunnel 1, a left tunnel 2, and a right tunnel 3. The left tunnel 2 is located between the horizontal pilot tunnel 1 and the right tunnel 3. An auxiliary construction passage 4 is provided between the left tunnel 2 and the horizontal pilot tunnel 1. A vehicle passage 5 and a pedestrian passage (not shown in the figure) are provided between the right tunnel 3 and the left tunnel 2. The openings of the left tunnel 2, the pedestrian passage, the remaining auxiliary construction passages 4 except the auxiliary construction passage 4 farthest from the opening, and the remaining vehicle passages 5 except the vehicle passage 5 farthest from the opening are closed.

[0056] The auxiliary construction passage 4 connects the horizontal pilot tunnel 1 and the left tunnel 2. The vehicle passage 5 connects the right tunnel 3 and the left tunnel 2. The pedestrian passage connects the right tunnel 3 and the left tunnel 2.

[0057] The auxiliary construction access 4 farthest from the tunnel entrance is the foremost auxiliary construction access 4 along the tunneling direction of the tunnel, and the vehicle access 5 farthest from the tunnel entrance is the foremost vehicle access 5 along the tunneling direction of the tunnel.

[0058] In this embodiment, the entrance of the left tunnel 2, the pedestrian access, the remaining auxiliary construction accesses 4 except the auxiliary construction access 4 farthest from the tunnel entrance, and the remaining vehicle accesses 5 except the vehicle access 5 farthest from the tunnel entrance are closed to form a negative pressure system inside the tunnel.

[0059] The hybrid ventilation system includes a first axial flow fan 6, a second axial flow fan 7, a third axial flow fan 8, an exhaust device 9, a detection device, and a control device.

[0060] The first axial flow fan 6 is arranged in the pilot tunnel 1, the second axial flow fan 7 and the third axial flow fan 8 are arranged in the right tunnel 3, and the exhaust device 9 is arranged in the left tunnel 2. The first axial flow fan 6 is used to send fresh air to the face of the pilot tunnel 1 through the air duct 10, the second axial flow fan 7 is used to send fresh air to the face of the left tunnel 2 through the air duct 10, the third axial flow fan 8 is used to send fresh air to the face of the right tunnel 3 through the air duct 10, and the exhaust device 9 is used to send the polluted air in the left tunnel 2 to the outside of the tunnel to form a circulating air flow with the pilot tunnel 1 and the right tunnel 3 for intake and the left tunnel 2 for exhaust.

[0061] The face, also known as the heading face, is a term in tunnel construction, referring to the working face where the tunnel (in coal mining, mining or tunnel engineering) is continuously advanced.

[0062] In this embodiment, the first axial flow fan 6 sucks the fresh air outside the tunnel into the pilot tunnel 1, the second axial flow fan 7 and the third axial flow fan 8 suck the fresh air outside the tunnel into the right tunnel 3, and the exhaust device 9 sucks the air in the left tunnel 2 out of the tunnel, creating a pressure difference between the left tunnel 2 and the pilot tunnel 1, and between the left tunnel 2 and the right tunnel 3, thus forming a negative pressure ventilation system in the three tunnels. The first axial flow fan 6 sends the fresh air to the face of the pilot tunnel 1 through the air duct 10 and simultaneously forms a return air flow. The return air flow of the pilot tunnel 1 enters the left tunnel 2 under the action of the pressure difference when passing through the first auxiliary construction access; the second axial flow fan 7 sends the fresh air to the face of the left tunnel 2 through the air duct 10 and simultaneously forms a return air flow; the third axial flow fan 8 sends the fresh air to the face of the right tunnel 3 and simultaneously forms a return air flow. The return air flow of the right tunnel 3 enters the left tunnel 2 under the action of the pressure difference when passing through the first vehicle access 5.

[0063] For example, the air outlet of the first axial flow fan 6 is connected to one end of a ventilation duct 10, and the other end of the ventilation duct 10 extends to a position about 15 meters away from the face of the pilot tunnel 1, so as to supply air to the face of the pilot tunnel 1; the air outlet of the second axial flow fan 7 is connected to one end of a ventilation duct 10, and the other end of the ventilation duct 10 extends to a position about 15 meters away from the face of the left tunnel 2 after passing through the first vehicle passage 5, so as to supply air to the face of the left tunnel 2; the air outlet of the third axial flow fan 8 is connected to one end of a ventilation duct 10, and the other end of the ventilation duct 10 extends to a position about 15 meters away from the face of the right tunnel 3, so as to supply air to the face of the right tunnel 3.

[0064] The number of detection devices is multiple, and one detection device is provided in each of the pilot tunnel 1, the left tunnel 2 and the right tunnel 3. The detection device in the pilot tunnel 1 is used to detect the air parameters in the pilot tunnel 1, the detection device in the left tunnel 2 is used to detect the air parameters in the left tunnel 2, and the detection device in the right tunnel 3 is used to detect the air parameters in the right tunnel 3.

[0065] The control device is respectively connected to the first axial flow fan 6, the second axial flow fan 7, the third axial flow fan 8, the exhaust device 9 and the detection device. The control device is used to control the first axial flow fan 6, the second axial flow fan 7, the third axial flow fan 8 and the exhaust device 9 according to the air parameters detected by the detection device.

[0066] In some embodiments of this embodiment, the first axial flow fan 6 is located between the opening of the first auxiliary construction passage and the entrance of the pilot tunnel 1, the second axial flow fan 7 is located between the opening of the first vehicle passage 5 and the entrance of the right tunnel 3, and the third axial flow fan 8 is located between the opening of the first vehicle passage 5 and the entrance of the right tunnel 3; wherein, the first auxiliary construction passage is the auxiliary construction passage farthest from the entrance of the pilot tunnel 1, and the first vehicle passage 5 is the vehicle passage 5 farthest from the entrance of the right tunnel 3.

[0067] In these embodiments, the first axial flow fan 6 is arranged between the opening of the first auxiliary construction passage and the entrance of the pilot tunnel 1, and the second axial flow fan 7 and the third axial flow fan 8 are arranged between the opening of the first vehicle passage 5 and the entrance of the right tunnel 3, so that the recirculated air in the pilot tunnel 1 can enter the left tunnel 2 under the action of the pressure difference before reaching the position of the first axial flow fan 6, and the recirculated air in the right tunnel 3 can enter the left tunnel 2 under the action of the pressure difference before reaching the positions of the second axial flow fan 7 and the third axial flow fan 8, which can effectively prevent the recirculated air from being sent back to the face by the axial flow fan.

[0068] For example, the horizontal pilot tunnel 1 has two side walls and a top wall. An auxiliary construction passage is provided on the first side wall of the horizontal pilot tunnel 1. The first axial flow fan 6 is arranged on the ground below the second side wall of the horizontal pilot tunnel 1 through a bracket. The distance between the first axial flow fan 6 and the projection of the first auxiliary construction passage on the second side wall of the horizontal pilot tunnel 1 is about 30 meters. The right tunnel 3 has two side walls and a top wall. A vehicle passage 5 is provided on the first side wall of the right tunnel 3. The second axial flow fan 7 is arranged on the ground below the first side wall of the right tunnel 3 through a bracket. The third axial flow fan 8 is arranged on the ground below the second side wall of the right tunnel 3 through a bracket. The distance between the second axial flow fan 7 and the opening of the first vehicle passage 5 on the first side wall of the right tunnel 3 is about 30 meters. The distance between the third axial flow fan 8 and the projection of the first vehicle passage 5 on the second side wall of the right tunnel 3 is about 30 meters.

[0069] In some embodiments of the present embodiment, a sealing wall is provided at the entrance of the left tunnel 2, and the sealing wall closes the entrance of the left tunnel 2; the exhaust device 9 is installed on the sealing wall, and the exhaust device 9 extracts the polluted air in the left tunnel 2 from the tunnel.

[0070] A safety evacuation door is provided on the sealing wall for evacuation in case of an emergency.

[0071] In some embodiments of the present embodiment, the exhaust device 9 includes a first exhaust fan and a second exhaust fan, and the first exhaust fan and the second exhaust fan are backup fans for each other.

[0072] For example, each time the first exhaust fan or the second exhaust fan is turned on; when the first exhaust fan is turned on, the second exhaust fan serves as a backup fan; when the second exhaust fan is turned on, the first exhaust fan serves as a backup fan.

[0073] In these embodiments, the first exhaust fan and the second exhaust fan are used as backup fans for each other. When one of the exhaust fans fails, the ventilation of the tunnel can still be carried out, improving the reliability.

[0074] In some embodiments of the present embodiment, the detection device includes a wind speed sensor, a temperature sensor, a negative pressure sensor, a carbon monoxide sensor, a carbon dioxide sensor, an oxygen concentration sensor, and a hydrogen sulfide sensor. The temperature sensor, the carbon monoxide sensor, the carbon dioxide sensor, the oxygen concentration sensor, and the hydrogen sulfide sensor are arranged at the heading face. The wind speed sensor is arranged at the air outlet of the air duct 10. The negative pressure sensor is arranged at the middle position between the heading face and the corresponding tunnel entrance.

[0075] For example, for the detection device arranged in the horizontal pilot tunnel 1, its temperature sensor, carbon monoxide sensor, carbon dioxide sensor, oxygen concentration sensor and hydrogen sulfide sensor are arranged at the heading face of the horizontal pilot tunnel 1, the wind speed sensor is arranged at the air outlet of the air duct 10 corresponding to the first axial flow fan 6, and the negative pressure sensor is arranged at the middle position between the heading face of the horizontal pilot tunnel 1 and the entrance of the horizontal pilot tunnel 1. For the detection device arranged in the left tunnel 2, its temperature sensor, carbon monoxide sensor, carbon dioxide sensor, oxygen concentration sensor and hydrogen sulfide sensor are arranged at the heading face of the left tunnel 2, the wind speed sensor is arranged at the air outlet of the air duct 10 corresponding to the second axial flow fan 7, and the negative pressure sensor is arranged at the middle position between the heading face of the left tunnel 2 and the entrance of the left tunnel 2. For the detection device arranged in the right tunnel 3, its temperature sensor, carbon monoxide sensor, carbon dioxide sensor, oxygen concentration sensor and hydrogen sulfide sensor are arranged at the heading face of the right tunnel 3, the wind speed sensor is arranged at the air outlet of the air duct 10 corresponding to the third axial flow fan 8, and the negative pressure sensor is arranged at the middle position between the heading face of the right tunnel 3 and the entrance of the right tunnel 3.

[0076] In some embodiments of this embodiment, the control device includes a data calculation unit, a fan control unit and a risk alarm unit. The data calculation unit is used to analyze the historical monitoring data and real-time monitoring data collected by the detection device by using a machine learning algorithm. The fan control unit is used to adjust the operating parameters of the first axial flow fan 6, the second axial flow fan 7, the third axial flow fan 8 and the exhaust device 9 according to the analysis result of the data calculation unit. The risk alarm unit is used to give a risk warning according to the detection result of the detection device.

[0077] In some embodiments of this embodiment, analyzing the historical monitoring data and real-time monitoring data collected by the detection device by using a machine learning algorithm includes: performing data preprocessing on the real-time monitoring data to obtain a time series matrix; inputting the time series matrix into a pre-trained hybrid neural network model to obtain the predicted values of the air volume demand index Q d , the negative pressure demand index P d and the fan speed demand index N d ; wherein, the method of data preprocessing includes: removing the values exceeding the preset standard range in the monitoring data as outliers. The monitoring data includes the wind speed v, the temperature T, the negative pressure P, the carbon monoxide concentration C CO , the carbon dioxide concentration the oxygen concentration and the hydrogen sulfide concentration For example, data that exceeds three standard deviations of the corresponding sensor range is regarded as outliers; the principal component analysis method is used to extract features and reduce the dimension of the monitoring data, and the original seven-dimensional data is converted into a three-dimensional principal component feature vector X = [x1, x2, x3]; a sliding time window data sequence is constructed. According to the standard of one window every 10 minutes, 60 data points in each window, and a monitoring frequency of once per second, the principal component feature vectors corresponding to these data points are arranged in sequence to form a time series matrix, which is used to capture the change trend of the data in the time dimension.

[0078] In some embodiments of this embodiment, the hybrid neural network model adopts a structure combining a convolutional neural network and a long short-term memory network. The hybrid neural network model includes a convolutional neural network, a long short-term memory network, and an output network.

[0079] The convolutional neural network includes two convolutional layers and a max pooling layer. The input of the convolutional neural network is the time series matrix; the convolutional kernel size of the first convolutional layer is 3×3, the stride is 1, the number of channels is 16, and the ReLU activation function is adopted; the convolutional kernel size of the second convolutional layer is 3×3, the stride is 1, the number of channels is 32, and the ReLU activation function is adopted; the pooling window size of the max pooling layer is 2×2, and the max pooling layer is used to extract the local features of the data and reduce the dimension.

[0080] The long short-term memory network includes two layers, and each layer has 64 hidden units. The input of the long short-term memory network is the feature sequence output by the convolutional neural network.

[0081] The output network includes an output layer and a fully connected layer before the output layer; the output layer has three nodes, which respectively correspond to the air volume demand index Q d , the negative pressure demand index P d and the fan speed demand index N d ; the number of nodes in the fully connected layer is 128, and the ReLU activation function is adopted. In these embodiments, by adding a fully connected layer in front of the output layer, the expression ability of the model is enhanced.

[0082] In some embodiments of this embodiment, the training method of the hybrid neural network model includes: performing data preprocessing on the historical monitoring data to obtain a time series matrix; dividing the time series matrix into a training set, a validation set, and a test set; setting the total number of training cycles, and in each training cycle, taking out the time series matrix sample data from the training set one by one and inputting it into the hybrid neural network model to obtain a predicted output value y i , corresponding to the air volume demand index Q of the fan control d , the negative pressure demand index P d and the fan speed demand index N d .

[0083] In some embodiments of the present embodiment, the operating parameters of the first axial flow fan 6, the second axial flow fan 7, the third axial flow fan 8 and the exhaust device 9 are adjusted according to the analysis results of the data calculation unit, including:

[0084] According to the change rate ΔC of the carbon monoxide concentration detected by the detection device corresponding to the axial flow fan CO and the air volume demand index Q d to determine the cooperative control coefficient α g ; if the change rate ΔC of the carbon monoxide concentration detected by the detection device corresponding to the axial flow fan CO is high and the air volume demand index Q d is large, then the cooperative control coefficient α g increases;

[0085] Calculate the target speeds N of the first axial flow fan 6, the second axial flow fan 7 and the third axial flow fan 8 target , and the calculation formula is:

[0086]

[0087] where k1 is the rotational speed - air volume proportionality coefficient of the axial flow fan, which is obtained from the model parameters of the axial flow fan;

[0088] Perform adaptive dynamic adjustment on the first exhaust fan and the second exhaust fan. The air volume Q of the first exhaust fan and the second exhaust fan is related to the blade angle θ, the rotational speed n and the air density ρ in the tunnel. The air density ρ is calculated according to the temperature T and negative pressure P in the tunnel, and the calculation formula is:

[0089]

[0090] where P0 is the standard physical atmospheric pressure and T0 is the absolute temperature under standard conditions;

[0091] The calculation formula for the air volume Q is:

[0092]

[0093] where k2 is a fixed coefficient and f(θ) is a function of the blade angle;

[0094] According to the air volume demand index Q d , the negative pressure demand index P d , the air density ρ and the performance curve of the exhaust fan, the particle swarm optimization algorithm is used to obtain the optimal rotational speed n target and the blade angle θ target .

[0095] In some embodiments of the present embodiment, the carbon monoxide concentration C is stored in the risk alarm unitCO 、 Carbon dioxide concentration and hydrogen sulfide concentration and other parameter alarm thresholds. When the values detected by the sensors exceed the corresponding alarm thresholds, alarm signals are sent to alert the staff.

[0096] The second aspect of this embodiment discloses a construction method, and the construction method includes step S100 and step S200.

[0097] Step S100. Pressurized ventilation is carried out when the driving distance is less than the first preset value.

[0098] As Figure 2 shown, the method of pressurized ventilation is: a first axial flow fan 6 is arranged at the entrance of the pilot tunnel 1, and the first axial flow fan 6 sends fresh air to the face of the pilot tunnel 1 through the air duct 10; a second axial flow fan 7 is arranged at the entrance of the left tunnel 2, and the second axial flow fan 7 sends fresh air to the face of the left tunnel 2 through the air duct 10; a third axial flow fan 8 is arranged at the entrance of the right tunnel 3, and the third axial flow fan 8 sends fresh air to the face of the right tunnel 3 through the air duct 10.

[0099] Step S200. Hybrid ventilation is carried out when the driving distance is greater than or equal to the first preset value.

[0100] The method of hybrid ventilation is: the first axial flow fan 6, the second axial flow fan 7, the third axial flow fan 8, the exhaust device 9, the detection device and the control device are arranged according to the hybrid ventilation system described in the first aspect of this embodiment; the entrance of the left tunnel 2, all pedestrian passages, all the remaining auxiliary construction passages 4 except the auxiliary construction passage 4 farthest from the entrance, and all the remaining vehicle passages 5 except the vehicle passage 5 farthest from the entrance are closed; wherein, every time the length of the second preset value is advanced forward, the first axial flow fan 6, the second axial flow fan 7, the third axial flow fan 8 and the detection device are moved forward a preset distance along the driving direction.

[0101] In this embodiment, the ventilation method is determined according to the driving distance. Taking the entrance as the reference point, pressurized ventilation is adopted when the driving distance is less than the first preset value, and hybrid ventilation is adopted when the driving distance is greater than or equal to the first preset value, so as to adapt to the ventilation in each stage of the tunnel excavation process.

[0102] In this embodiment, the first preset value and the second preset value, as well as the distance that the first axial flow fan 6, the second axial flow fan 7, the third axial flow fan 8 and the detection device move forward each time, are determined according to the settings of the auxiliary construction passage 4 and the vehicle passage 5 in the tunnel, and the performance of equipment such as the axial flow fans and the detection device.

[0103] For example, for a certain tunnel, a vehicle passage 5 is provided every 750 m, and an auxiliary construction passage 4 is provided every 1500 m. Then, when the tunneling distance is less than 1600 m, forced ventilation is adopted; when the tunneling distance is greater than or equal to 1600 m, hybrid ventilation is adopted. After that, the first axial flow fan 6, the second axial flow fan 7, the third axial flow fan 8 and the detection device are moved every 1500 m of tunneling.

[0104] In this embodiment, the first axial flow fan 6, the second axial flow fan 7, the third axial flow fan 8 and the detection device move forward continuously as the tunnel advances. When applied to extra-long tunnels, jet fans do not need to be added on the premise of ensuring the ventilation effect, and the demand for the air duct 10 is reduced, greatly reducing the ventilation cost of extra-long tunnel construction.

[0105] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; and these modifications or replacements will not cause the essence of the corresponding technical solutions to deviate from the protection scope of the technical solutions of the embodiments of the present invention.

Claims

1. A hybrid ventilation system combining exhaust and pressure in extra-long tunnels, characterized in that, The extra-long tunnel includes a pilot tunnel, a left tunnel, and a right tunnel. An auxiliary construction passage is provided between the left tunnel and the pilot tunnel, and a vehicle passage and a pedestrian passage are provided between the right tunnel and the left tunnel. The entrances of the left tunnel, the pedestrian passage, the remaining auxiliary construction passages except the one farthest from the entrance of the tunnel, and the remaining vehicle passages except the one farthest from the entrance of the tunnel are closed. The hybrid ventilation system includes a first axial flow fan installed in the pilot tunnel, a second axial flow fan and a third axial flow fan installed in the right tunnel, an exhaust device, a detection device, and a control device installed in the left tunnel. Detection devices are installed in the pilot tunnel, the left tunnel, and the right tunnel. Among them, the first axial flow fan is used to supply fresh air to the face of the pilot tunnel through a ventilation duct, the second axial flow fan is used to supply fresh air to the face of the left tunnel through a ventilation duct, the third axial flow fan is used to supply fresh air to the face of the right tunnel through a ventilation duct, and the exhaust device is used to discharge the polluted air in the left tunnel out of the tunnel, so as to form a circulating air flow with the pilot tunnel and the right tunnel supplying air and the left tunnel exhausting air. Among them, the detection device is used to detect the air parameters in the pilot tunnel, the left tunnel, and the right tunnel, and the control device is used to control the first axial flow fan, the second axial flow fan, the third axial flow fan, and the exhaust device according to the detected air parameters.

2. The hybrid ventilation system of the extra-long tunnel combined with pumping and pressure according to claim 1 is characterized in that: The first axial flow fan is located between the opening of the first construction auxiliary passage and the entrance of the pilot tunnel, the second axial flow fan is located between the opening of the first vehicle passage and the entrance of the right tunnel, and the third axial flow fan is located between the opening of the first vehicle passage and the entrance of the right tunnel; among them, the first construction auxiliary passage is the construction auxiliary passage farthest from the entrance of the pilot tunnel, and the first vehicle passage is the vehicle passage farthest from the entrance of the right tunnel.

3. The combined extraction and pressure ventilation system of the extra-long tunnel according to claim 1 is characterized in that: The exhaust device includes a first exhaust fan and a second exhaust fan, and the first exhaust fan and the second exhaust fan are backup fans for each other.

4. The hybrid ventilation system combining exhaust and pressure in an extra-long tunnel according to claim 1, characterized in that, The detection device includes an air velocity sensor, a temperature sensor, a negative pressure sensor, a carbon monoxide sensor, a carbon dioxide sensor, an oxygen concentration sensor, and a hydrogen sulfide sensor. The temperature sensor, the carbon monoxide sensor, the carbon dioxide sensor, the oxygen concentration sensor, and the hydrogen sulfide sensor are installed at the face, the air velocity sensor is installed at the air outlet of the ventilation duct, and the negative pressure sensor is installed at the middle position between the face and the corresponding entrance of the tunnel.

5. The hybrid ventilation system with combined extraction and pressure in extra-long tunnels according to claim 1, characterized in that The control device includes a data calculation unit, a fan control unit, and a risk alarm unit. The data calculation unit is used to analyze the historical monitoring data and real-time monitoring data collected by the detection device using a machine learning algorithm. The fan control unit is used to adjust the operating parameters of the first axial flow fan, the second axial flow fan, the third axial flow fan, and the exhaust device according to the analysis results of the data calculation unit. The risk alarm unit is used to give a risk warning according to the detection results of the detection device.

6. The combined extraction and pressure ventilation system of the extra-long tunnel according to claim 5 is characterized in that: Analyzing the historical monitoring data and real-time monitoring data collected by the detection device using a machine learning algorithm includes: Preprocess the real-time monitoring data to obtain a time series matrix; input the time series matrix into a pre-trained hybrid neural network model to obtain the predicted values of the air volume demand index Q d , the negative pressure demand index P d and the fan speed demand index N d . Among them, the data preprocessing methods include: removing the values in the monitoring data that exceed the preset standard range, and the monitoring data includes wind speed v, temperature T, negative pressure P, carbon monoxide concentration C CO , carbon dioxide concentration oxygen concentration and hydrogen sulfide concentration Feature extraction and dimensionality reduction are performed on the monitoring data using the principal component analysis method, and the original seven-dimensional data is converted into a three-dimensional principal component feature vector X = [x1, x2, x3]; a sliding time window data sequence is constructed. According to the standard of one window every 10 minutes, 60 data points in each window, and a monitoring frequency of once per second, the principal component feature vectors corresponding to these data points are arranged in sequence to form a time series matrix.

7. The hybrid ventilation system combining exhaust and pressure in a long tunnel according to claim 6, characterized in that, The hybrid neural network model includes a convolutional neural network, a long short-term memory network, and an output network. The convolutional neural network includes two convolutional layers and a max pooling layer. The input of the convolutional neural network is a time series matrix. The convolutional kernel size of the first convolutional layer is 3×3, the stride is 1, the number of channels is 16, and the ReLU activation function is used. The convolutional kernel size of the second convolutional layer is 3×3, the stride is 1, the number of channels is 32, and the ReLU activation function is used. The pooling window size of the max pooling layer is 2×2; The long short-term memory network includes two layers, and each layer has 64 hidden units. The input of the long short-term memory network is the feature sequence output by the convolutional neural network; The output network includes an output layer and a fully connected layer located before the output layer; the output layer has three nodes, and the three nodes respectively correspond to the air volume demand index Q for fan control d , the negative pressure demand index P d and the fan speed demand index N d ; the number of nodes in the fully connected layer is 128, and the ReLU activation function is adopted.

8. The combined extraction and pressure ventilation system of the extra-long tunnel according to claim 7 is characterized in that: The training method of the hybrid neural network model includes: Perform data preprocessing on the historical monitoring data to obtain a time series matrix; divide the time series matrix into a training set, a validation set, and a test set; Set the total number of training cycles. In each training cycle, sequentially extract the time series matrix sample data from the training set, input it into the hybrid neural network model, and obtain the predicted output value y i , corresponding to the air volume demand index Q for fan control d , negative pressure demand index P d and fan speed demand index N d .

9. The hybrid ventilation system combining exhaust and pressure in a long tunnel according to claim 6, wherein, Adjust the operating parameters of the first axial flow fan, the second axial flow fan, the third axial flow fan, and the exhaust device according to the analysis results of the data calculation unit, including: According to the change rate ΔC of the carbon monoxide concentration detected by the detection device corresponding to the axial flow fan CO and the air volume demand index Q d to determine the collaborative control coefficient α g ; Calculate the target rotational speeds N of the first axial flow fan, the second axial flow fan, and the third axial flow fan target , and the calculation formula is as follows: where k1 is the rotational speed - air volume ratio coefficient of the axial flow fan; Calculate the air density ρ according to the temperature T and negative pressure P in the tunnel, and the calculation formula is: where P0 is the standard physical atmospheric pressure and T0 is the absolute temperature under standard conditions; The calculation formula for the air volume Q is: where k2 is a fixed coefficient and f(θ) is a function of the blade angle; According to the air volume demand index Q d , the negative pressure demand index P d , the air density ρ, and the performance curve of the exhaust fan, the optimal rotational speed n target and the blade angle θ target are obtained by using the particle swarm optimization algorithm 10. A construction method, characterized in that, including: S100. Perform forced ventilation when the driving distance is less than the first preset value. The method of forced ventilation is: Set the first axial flow fan at the entrance of the pilot tunnel, and the first axial flow fan sends fresh air to the face of the pilot tunnel through the air duct; Set the second axial flow fan at the entrance of the left tunnel, and the second axial flow fan sends fresh air to the face of the left tunnel through the air duct; Set the third axial flow fan at the entrance of the right tunnel, and the third axial flow fan sends fresh air to the face of the right tunnel through the air duct; S200. Perform hybrid ventilation when the driving distance is greater than or equal to the first preset value. The method of hybrid ventilation is: Set the first axial flow fan, the second axial flow fan, the third axial flow fan, the exhaust device, the detection device, and the control device according to the hybrid ventilation system described in any one of claims 1-9; Seal the entrance of the left tunnel, all pedestrian passages, all auxiliary construction passages except the auxiliary construction passage farthest from the entrance, and all vehicle passages except the vehicle passage farthest from the entrance; Among them, Whenever the length of the forward driving reaches the second preset value, move the first axial flow fan, the second axial flow fan, the third axial flow fan, and the detection device forward a preset distance along the driving direction.

Images