Construction method and control method of non-frequency conversion fan control model for highway tunnel fire

Abstract

The present invention discloses a method for constructing a control model of a non-inverter fan for highway tunnel fires and a control method, including: First, multiple groups of fan control parameters of non-inverter fans under different fire characteristic parameters, tunnel structure facility parameters, and initial environmental parameters can be obtained, and a data set is constructed with these parameters as samples. Then, a control model is constructed with the fixed tunnel parameters, initial environmental parameters, smoke spread position, and target wind speed as input data and the fan control parameters as output data. Finally, the constructed data set is used to train the control model, so as to obtain a non-inverter fan control model that can calculate the fan control parameters of non-inverter fans in the tunnel according to the fixed tunnel parameters, initial environmental parameters, smoke spread position, and target wind speed. The fan control parameters in the tunnel are dynamically adjusted through the non-inverter fan control model to adapt to the fire change situation and realize the intelligent control of longitudinal smoke exhaust for tunnel fires.

Description

Technical Field

[0001] The present invention relates to the technical field of tunnel ventilation control, and particularly relates to a method for constructing a control model of a non-frequency conversion fan for highway tunnel fires and a control method. Background Art

[0002] The existing longitudinal smoke exhaust for tunnel fires mainly forms a fixed critical wind speed or sub-critical wind speed upstream of the fire source by starting the jet fans in the tunnel, and blows the fire smoke downstream of the fire source for discharge. Since the power of the fire source changes dynamically during a fire, the fixed critical wind speed or sub-critical wind speed cannot adapt to the change of the fire source, resulting in an excessive wind speed in the initial stage of the fire, which destroys the smoke stratification downstream of the fire source and is not conducive to the evacuation of people downstream of the fire source. Summary of the Invention

[0003] Aiming at the deficiencies of the existing technology, the present invention proposes a method for constructing a control model of a non-frequency conversion fan for highway tunnel fires and a control method, which can realize the intelligent control of longitudinal smoke exhaust for highway tunnel fires through non-frequency conversion fans. The specific technical solutions are as follows:

[0004] In the first aspect, a method for constructing a control model of a non-frequency conversion fan for highway tunnel fires is provided. In the first feasible implementation manner of the first aspect, it includes:

[0005] Obtain multiple sets of smoke exhaust data of non-frequency conversion fans to construct a data set. The smoke exhaust data includes tunnel fixed parameters, initial environmental parameters, smoke spread position, target wind speed, and fan control parameters;

[0006] Use the tunnel fixed parameters, initial environmental parameters, smoke spread position, and target wind speed as input data, and the fan control parameters as output data to construct a control model;

[0007] Train the control model through the data set to obtain a non-frequency conversion fan control model.

[0008] Combined with the first feasible implementation manner of the first aspect, in the second feasible implementation manner of the first aspect, obtaining multiple sets of smoke exhaust data of non-frequency conversion fans to construct a data set includes:

[0009] Use simulation software to construct a highway tunnel model, and conduct simulation tests through the highway tunnel model to obtain the smoke exhaust data of non-frequency conversion fans under different smoke exhaust conditions.

[0010] Combined with the first feasible implementation manner of the first aspect, in the third feasible implementation manner of the first aspect, obtaining multiple sets of smoke exhaust data of non-frequency conversion fans to construct a data set includes:

[0011] Calculate the relative error between the target wind speed and the control wind speed in the fan control parameters for each group of smoke exhaust data, and filter out the smoke exhaust data with a relative error less than the error threshold to construct the dataset.

[0012] Combined with the first implementation manner of the first aspect, in the fourth implementation manner of the first aspect, the tunnel fixed parameters include the tunnel length, the outlet wind speed, the fan outlet area, the number of fans, and / or the fan position.

[0013] Combined with the first implementation manner of the first aspect, in the fifth implementation manner of the first aspect, training the control model with the dataset includes:

[0014] Calculate the global error of the control model according to the control wind speed in the smoke exhaust data and the control wind speed output by the control model, and compare the global error with the set accuracy value;

[0015] In response to the global error being greater than the accuracy value, continue to train through the dataset.

[0016] Combined with the fifth implementation manner of the first aspect, in the sixth implementation manner of the first aspect, training the control model with the dataset includes:

[0017] In response to the global error being less than the accuracy value, compare the smoke spread position in the smoke exhaust data with the fan start position output by the model;

[0018] If there is no intersection between the smoke spread position and the fan start position, the control model training is completed; otherwise, continue to train the control model through the dataset.

[0019] In the second aspect, a non-inverter fan control method for highway tunnel fires is provided. In the first implementation manner of the second aspect, it includes:

[0020] Adopt the construction method described in any one of the first to sixth implementation manners of the first aspect to construct a non-inverter fan control model;

[0021] Obtain the tunnel fixed parameters, environmental initial parameters, smoke spread position, and target wind speed of the tunnel to be controlled, and determine the fan control parameters in the tunnel through the trained non-inverter fan control model.

[0022] Combined with the first implementation manner of the second aspect, in the second implementation manner of the second aspect, determining the fan control parameters through the non-inverter fan control model includes:

[0023] Verify the fan control parameters based on the smoke spread position. If the verification fails, eliminate the fan control parameters and re-determine the fan control parameters through the non-variable frequency fan control model.

[0024] Combined with the second implementation manner of the second aspect, in the third implementation manner of the second aspect, verifying the fan control parameters includes:

[0025] Compare the smoke spread position with the fan positions in the tunnel fixed parameters to determine the fan positions that intersect with the smoke spread position;

[0026] Based on the fan start positions and the number of fans started in the fan control parameters, determine whether there is a fan start among the fan positions that intersect with the smoke spread position;

[0027] In response to there being a fan start among the fan positions that intersect with the smoke spread position, the verification fails.

[0028] Combined with the second implementation manner of the second aspect, in the third implementation manner of the second aspect, verifying the fan control parameters includes:

[0029] If the verification passes, use the fan control parameters as alternative control parameters and continue to calculate the fan control parameters through the non-variable frequency fan control model until the number of alternative control parameters reaches the quantity threshold;

[0030] Select the fan control parameter corresponding to the minimum control wind speed from all the alternative control parameters.

[0031] Beneficial effects: By using the method for constructing a non-variable frequency fan control model and the control method for highway tunnel fires of the present invention, a machine learning model with tunnel fixed parameters, environmental initial parameters, smoke spread position, and target wind speed as input variables and the fan control parameters of non-variable frequency fans as output variables can be trained through a data set composed of the smoke exhaust data of non-variable frequency fans in different highway tunnel fire occurrence scenarios. The trained non-variable frequency fan control model can dynamically adjust the number of fans started, start positions, and control wind speed in the tunnel according to the tunnel fixed parameters, as well as the real-time smoke spread position and target wind speed at the fire site, so as to adapt to the fire change situation in the tunnel and achieve intelligent longitudinal smoke exhaust control for highway tunnel fires. Description of the Drawings

[0032] In order to more clearly illustrate the specific embodiments of the present invention, the drawings required for use in the specific embodiments will be briefly introduced below. In all the drawings, the components or parts do not necessarily draw according to the actual scale.

[0033] Figure 1Flowchart of the method for constructing a non-frequency conversion fan control model for highway tunnel fires provided by an embodiment of the present invention;

[0034] Figure 2 Flowchart of the non-frequency conversion fan control method for highway tunnel fires provided by an embodiment of the present invention;

[0035] Figure 3 Schematic diagram of the zoning of tunnel fans. Detailed implementation manners

[0036] Hereinafter, embodiments of the technical solution of the present invention will be described in detail with reference to the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, so they are only examples and cannot be used to limit the protection scope of the present invention.

[0037] As Figure 1 shown in the flowchart of the method for constructing a non-frequency conversion fan control model for highway tunnel fires, the construction method includes:

[0038] Step 1: Obtain multiple sets of smoke exhaust data of non-frequency conversion fans to construct a data set, and the smoke exhaust data includes tunnel fixed parameters, initial environmental parameters, smoke spread position, target wind speed, and fan control parameters;

[0039] Step 2: Use the tunnel fixed parameters, initial environmental parameters, smoke spread position, and target wind speed as input data, and the fan control parameters as output data to construct a control model;

[0040] Step 3: Train the control model through the data set to obtain a non-frequency conversion fan control model.

[0041] Specifically, first, multiple sets of fan control parameters and target wind speeds of non-frequency conversion fans under different fire characteristic parameters, tunnel structure facility parameters, and initial environmental parameters can be obtained, and these parameters are used as samples to construct a data set. Then, use the tunnel fixed parameters, initial environmental parameters, smoke spread position, and target wind speed as input data, and the fan control parameters as output data to construct a control model. Finally, train the control model through the constructed data set, so as to obtain a non-frequency conversion fan control model that can be used to calculate the fan control parameters of non-frequency conversion fans in the tunnel according to the tunnel fixed parameters, initial environmental parameters, smoke spread position, and target wind speed during a fire. To adapt to the fire change situation in the tunnel and realize the longitudinal smoke exhaust adaptive control of highway tunnel fires.

[0042] It should be understood that there are many indicators affecting the ventilation wind speed of highway tunnels, involving tunnel structure facility parameters, tunnel fire characteristics, and tunnel initial environmental parameters, etc.

[0043] The tunnel structure facility parameters affecting longitudinal smoke exhaust in case of fire mainly include the length and fan-related parameters. Among them, the frictional resistance generated by the tunnel length will reduce the tunnel wind speed. Parameters such as the wind speed at the fan outlet, the outlet area, the number of fans and the number of fans in operation, as well as the fan location, etc. are the basis for tunnel ventilation. The fan layout position is generally divided into three parts: the tunnel entrance area, the middle area, and the exit area. When a fire occurs in the tunnel, if a fire or smoke appears within 120m near the fan, the fan in this area cannot be turned on.

[0044] The tunnel fire characteristic parameters affecting longitudinal smoke exhaust in case of fire mainly include the location range of the fire occurrence and the smoke spread range. The different locations of the tunnel fire occurrence directly affect the fan turning-on positions in the tunnel. For fires of different scales, the amount of smoke generated and the speed of smoke spread are different. The smoke will spread along the tunnel top and side walls to the upstream and downstream of the tunnel, and the fans at the covered positions cannot be turned on. Therefore, when constructing the control model, the smoke spread position also needs to be considered.

[0045] The tunnel initial environment parameters affecting longitudinal smoke exhaust in case of fire mainly include the magnitude and direction of the wind speed. The greater the initial environment wind speed in the tunnel, the smaller the wind speed required for the controlled fan; the smaller the environment wind speed, the greater the wind speed required for the controlled fan. Therefore, when constructing the control model, the initial environment wind speed also needs to be considered.

[0046] Based on the above analysis, information such as the tunnel length, fan location, fan outlet wind speed, fan outlet area, number of fans, number of fans in operation, environment wind speed and direction, fire location range, and fire smoke spread range will all affect the control of the wind speed in the tunnel. Therefore, the smoke exhaust data can include the data corresponding to the tunnel length, fan location, fan outlet wind speed, fan outlet area, number of fans, number of fans in operation, environment wind speed and direction, fire location range, fire smoke spread position, and target wind speed.

[0047] In this embodiment, optionally, obtain multiple sets of smoke exhaust data of non-inverter fans to construct a data set, including:

[0048] Use simulation software to construct a highway tunnel model and conduct simulation tests through the highway tunnel model to obtain the smoke exhaust data of non-inverter fans under different smoke exhaust conditions.

[0049] Specifically, since it is difficult to conduct on-site experimental research on the actual tunnel fan control. For this reason, Fluent software can be used to construct a highway tunnel model, and the critical wind speed or sub-critical wind speed required to ensure the safe evacuation and escape of trapped personnel upstream and downstream of the tunnel can be used as the target wind speed for longitudinal smoke exhaust control in highway tunnel fires, and tunnel ventilation simulation can be carried out.

[0050] Such as Figure 3As shown in the figure, considering the influence of the location of tunnel fire on the starting position of the fan facilities, the tunnel is divided into 7 regions according to the location of the fans in the tunnel. Tunnel fire smoke and spread When in the inlet fan area, the middle fan area or the outlet fan area, the fans in this area cannot be turned on. The inlet fan area is , the middle fan area is , the outlet fan area is , , are the coordinates of the first fan and the last fan at the inlet position. , are the coordinates of the first fan and the last fan at the middle position; and are the coordinates of the first fan and the last fan at the outlet position respectively. Among them, is the fire occurrence position, , are the spread lengths of the flame smoke in the upstream and downstream of the tunnel respectively, represents the number of tunnel fires.

[0051] , , are the numbers of fans in the inlet fan area, the middle fan area and the outlet fan area respectively.

[0052] Through the simulation analysis of the constructed highway tunnel model, the control wind speed in the highway tunnel can be obtained under different tunnel lengths, fan positions, fan outlet wind speeds, fan outlet areas, fan numbers, fan opening numbers, and environmental initial wind speed and direction parameters. Taking the tunnel length, fan position, fan outlet wind speed, fan outlet area, fan number, fan opening number, environmental wind speed and direction, fire position range, and fire smoke spread position as samples, a data set is constructed as shown in the following table:

[0053]

[0054] Among them, the number of fans in the tunnel should be greater than or equal to the number of fans turned on, that is ; in the parameters are the numbers of fans starting in the inlet, middle and outlet areas respectively, satisfying , and at the same time, the number of fans in each area satisfies , , .

[0055] In this embodiment, optionally, a data set is constructed by obtaining the smoke exhaust data of multiple non-variable frequency fans, including: calculating the relative error between the target wind speed and the control wind speed in the control parameters of each group of smoke exhaust data respectively, and screening out the smoke exhaust data with a relative error less than the error threshold to construct the data set.

[0056] Specifically, when constructing the data set, the target wind speed and the control wind speed in the smoke exhaust data obtained from each simulation analysis can be compared, and the relative error between the target wind speed and the control wind speed can be calculated. If the relative error between the target wind speed and the control wind speed exceeds the error threshold, it indicates that the current simulation does not achieve good smoke exhaust control. Therefore, the smoke exhaust data obtained from the current simulation can be excluded and not included in the data set to avoid affecting the model training accuracy. The specific calculation formula for the relative error is as follows:

[0057] 。

[0058] Among them, 、 are the target wind speed and the control wind speed obtained from the simulation analysis respectively.

[0059] In this embodiment, a BP neural network can be used as the control model. The input parameters include tunnel length, outlet wind speed, fan outlet area, number of fans, fan position, initial environmental wind speed, fire smoke spread position, and target wind speed. The output parameters include control wind speed, number of started fans, and fan start positions.

[0060] During training, first initialize the weights of each node, the error function e, the calculation accuracy value, and the maximum number of learning times M, and determine the number of hidden neurons , and the specific calculation formula is as follows:

[0061] ;

[0062] Among them, is the number of neurons in the input layer, is the number of neurons in the output layer, is a constant between 0 and 10. Then use the data set to train the BP neural network, and calculate the outputs and of the hidden layer and the output layer in sequence. Their calculation expressions are:

[0063] , ;

[0064] ,

[0065] ;

[0066] Among them, is the input data in the sample data, is the weight of the input layer, is the input data in the is the bias of the input layer, is the weight of the hidden layer, is the input of the hidden layer, is the bias of the hidden layer. is the output parameter for controlling the model, including the control wind speed, the number of started fans, and the starting positions of the fans.

[0067] In this embodiment, optionally, training the control model by using the data set includes:

[0068] Calculating the global error of the control model according to the control wind speed in the smoke exhaust data and the control wind speed output by the control model, and comparing the global error with a set precision value;

[0069] When the global error is greater than the precision value, the gradient descent method is used to update the weight and bias data in the neural network, and training is continued from the samples through the data set again.

[0070] Specifically, during training, the actual control wind speed can be compared with the control wind speed output by the model, and the global error E of the error function is calculated. The specific calculation formula is as follows:

[0071] ;

[0072] Wherein, is the number of samples in the data set, is the number of output samples. In the present invention, the number is taken as the constant 3, is the actual value of the control wind speed in the sample, is the value of the control wind speed output by the model.

[0073] As Figure 3 shown, the fan layout positions are generally divided into three parts: the tunnel entrance area, the middle area, and the exit area. Since a fire or smoke appears within 120 m near the fan when a tunnel fire occurs, the fan in this area cannot be started. Therefore, to prevent the smoke from spreading to these three fan position areas, when the global error is less than the precision value, it is also necessary to compare the relationship between the smoke spreading position in the sample and the fan position. If the smoke spreading position intersects with any one of the three fan position areas divided, and there is a started fan in the intersecting fan position area, continue to extract samples from the data set for training. That is:

[0074] ① When When the smoke spread position input to the model intersects with the tunnel entrance area, and the fan start positions output by the model include the fans within the entrance area, continue to extract samples for training;

[0075] ② When When the smoke spread position input to the model intersects with the middle area of the tunnel, and the fan start positions output by the model include the fans within the middle area, continue to extract samples for training;

[0076] ③ When When the smoke spread position input to the model intersects with the tunnel exit area, and the fan start positions output by the model include the fans within the exit area, continue to extract samples for training.

[0077] Conversely, if the smoke spread position in the sample does not intersect with any of the three fan position areas divided, and there is no fan start in the fan position area where the fire intersects, it indicates that the model training is completed, and the non-inverter fan control model is obtained.

[0078] When a fire occurs in the tunnel, the initial environmental wind speed and the fire smoke spread position can be obtained in real time through the sensors in the tunnel, and the target wind speed can be determined by comparing with the pre-constructed database according to the initial environmental wind speed, the fire smoke spread position and the tunnel fixed parameters.

[0079] Combining the initial environmental wind speed, the fire smoke spread position, the tunnel fixed parameters and the target wind speed, determine the number of fans to be started, the fan start positions and the control wind speed in the tunnel through the trained non-inverter fan control model, so as to control the opening of the fans at the corresponding positions in the tunnel to adapt to the fire change situation in the tunnel, realize the longitudinal smoke exhaust adaptive control of highway tunnel fires, and avoid the interaction between fans at the same time.

[0080] As Figure 2 shown in the flowchart of the non-inverter fan control method for highway tunnel fires, the control method includes:

[0081] Step S1: Construct a non-inverter fan control model by using the above construction method;

[0082] Step S2: Obtain the tunnel fixed parameters, the initial environmental parameters, the smoke spread position and the target wind speed of the tunnel to be controlled, and determine the fan control parameters in the tunnel through the trained non-inverter fan control model.

[0083] Specifically, first, the non-inverter fan control model can be constructed using the above construction method. Then, when a fire occurs in the tunnel, the initial environmental wind speed and the location of the spread of fire smoke can be obtained in real time through the sensors in the tunnel. Based on the initial environmental wind speed, the location of the spread of fire smoke, and the fixed tunnel parameters, the target wind speed can be determined by comparing with the pre-constructed database. Combining the initial environmental wind speed, the location of the spread of fire smoke, the fixed tunnel parameters, and the target wind speed, and inputting the initial environmental wind speed, the location of the spread of fire smoke, the fixed tunnel parameters, and the target wind speed into the trained non-inverter fan control model, the number of fans to be started, the fan start positions, and the control wind speed in the tunnel can be determined through the trained non-inverter fan control model, so as to control the opening of the fans at the corresponding positions in the tunnel to adapt to the fire changes in the tunnel, achieve the longitudinal smoke exhaust adaptive control of highway tunnel fires, and avoid the interaction between fans at the same time.

[0084] In this embodiment, optionally, determining the fan control parameters through the non-inverter fan control model includes:

[0085] Verifying the fan control parameters through the smoke spread position. If the verification fails, the fan control parameters are excluded, and the fan control parameters are determined again through the non-inverter fan control model.

[0086] Specifically, when a tunnel fire occurs, if a fire or smoke appears within 120m near the fan, the fan in this area cannot be started. Therefore, after calculating the number of fans to be started, the fan start positions, and the control wind speed required currently through the non-inverter fan control model, it is also necessary to compare the current smoke spread position with different fan positions to determine whether the smoke spread position intersects with the fan start positions. If the smoke spread position intersects with the fan start positions, the fan control parameters output by the model cannot be used to control the non-inverter fans in the tunnel to avoid incorrect operations.

[0087] In this embodiment, optionally, verifying the fan control parameters includes:

[0088] Comparing the smoke spread position with the fan positions in the fixed tunnel parameters to determine the fan positions that intersect with the smoke spread position;

[0089] Based on the fan start positions and the number of fans to be started in the fan control parameters, determining whether there is a fan start among the fan positions that intersect with the smoke spread position;

[0090] In response to there being a fan start among the fan positions that intersect with the smoke spread position, the verification fails.

[0091] Specifically, when verifying the fan control parameters, first compare the smoke spread position with the fan positions in the tunnel fixed parameters to determine the fan positions intersecting with the smoke spread position. Then, based on the fan start positions and the number of started fans in the fan control parameters, determine whether the number of started fans at the fan positions intersecting with the smoke spread position is 0. If it is, the fan control parameter is available; otherwise, it is not available.

[0092] For example, when the smoke spread position intersects with the tunnel entrance area, if the number of started fans at the fan start positions in the fan start positions output by the model is 0, then take the control wind speed, the number of started fans, and the fan start positions as a set of alternative smoke exhaust data and store them in the alternative database; otherwise, re-obtain the initial environmental parameters, the smoke spread position, and the target wind speed and continue the calculation.

[0093] When the smoke spread position intersects with the middle area, if the number of started fans at the fan start positions in the fan start positions output by the model is 0, then take the control wind speed, the number of started fans, and the fan start positions as a set of alternative smoke exhaust data and store them in the alternative database; otherwise, re-obtain the initial environmental parameters, the smoke spread position, and the target wind speed and continue the calculation.

[0094] When the smoke spread position intersects with the tunnel exit area, if the number of started fans at the fan start positions in the fan start positions output by the model is 0, then take the control wind speed, the number of started fans, and the fan start positions as a set of alternative smoke exhaust data and store them in the alternative database; otherwise, re-obtain the initial environmental parameters, the smoke spread position, and the target wind speed and continue the calculation.

[0095] In this embodiment, optionally, verifying the fan control parameters includes:

[0096] If the verification is passed, take the fan control parameter as an alternative control parameter and continue to calculate the fan control parameter through the non-variable frequency fan control model until the number of alternative control parameters reaches the number threshold;

[0097] Select the fan control parameter corresponding to the minimum control wind speed from all the alternative control parameters.

[0098] Specifically, to improve the accuracy of non-inverter fan control, if the position where the smoke spreads does not intersect with the above three fan areas, the fan control parameters output by the model can be directly used as alternatives and stored in the alternative database. Then, the initial environmental parameters, the position where the smoke spreads, and the target wind speed are re-acquired, and the fan control parameters are continuously calculated through the non-inverter fan control model until the number of fan control parameters stored in the alternative database exceeds the set number threshold.

[0099] Finally, the control wind speeds of all the fan control parameters stored in the alternative database can be compared, and the fan control parameter with the minimum control wind speed can be selected as the optimal control parameter to control the non-inverter fans in the tunnel.

[0100] 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 or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered by the scope of the claims and the specification of the present invention.

Claims

1. A method for constructing a non-variable frequency fan control model for a highway tunnel fire, characterized in that: include: Acquire multiple sets of smoke exhaust data from non-variable frequency fans to build a data set. The smoke exhaust data includes tunnel fixed parameters, environmental initial parameters, smoke spreading location, target wind speed, and fan control parameters. A control model is constructed using the tunnel fixed parameters, environmental initial parameters, smoke spreading position, and target wind speed as input data and the fan control parameters as output data; The control model is trained by using the data set to obtain a non-variable frequency fan control model; The control model is trained using the data set, including: Calculating a global error of the control model according to the control wind speed in the smoke exhaust data and the control wind speed output by the control model, and comparing the global error with a set accuracy value; In response to the global error being greater than the precision value, continuing training with the data set; In response to the global error being less than the accuracy value, comparing the smoke spreading position in the smoke exhaust data with the fan start position output by the model; If there is no intersection between the smoke spreading position and the fan start position, the control model training is completed. Otherwise, the control model training continues through the data set.

2. The method for constructing a non-variable frequency fan control model for a highway tunnel fire according to claim 1 is characterized in that: Obtain multiple sets of smoke exhaust data from non-variable frequency fans to build a data set, including: A highway tunnel model was constructed using simulation software, and simulation experiments were carried out on the highway tunnel model to obtain smoke exhaust data of non-variable frequency fans under different smoke exhaust conditions.

3. The method for constructing a non-variable frequency fan control model for a highway tunnel fire according to claim 1, characterized in that: Obtain multiple sets of smoke exhaust data from non-variable frequency fans to build a data set, including: The relative error between the target wind speed in each set of smoke exhaust data and the controlled wind speed in the fan control parameter is calculated respectively, and the smoke exhaust data with a relative error less than an error threshold are screened out to construct the data set.

4. The method for constructing a non-variable frequency fan control model for a highway tunnel fire according to claim 1, characterized in that: The fixed parameters of the tunnel include tunnel length, outlet wind speed, fan outlet area, number of fans and / or fan positions.

5. A method for controlling a non-variable frequency fan in a highway tunnel fire, characterized in that: include: Using the construction method as described in any one of claims 1 to 4, a non-variable frequency fan control model is constructed; The fixed parameters of the controlled tunnel, the initial parameters of the environment, the smoke spreading position and the target wind speed are obtained, and the fan control parameters in the tunnel are determined through the trained non-variable frequency fan control model.

6. The method for controlling a non-variable frequency fan for a highway tunnel fire according to claim 5, characterized in that: The fan control parameters are determined by the non-variable frequency fan control model, including: The fan control parameters are verified through the smoke spreading position. If the verification fails, the fan control parameters are eliminated and the fan control parameters are re-determined through a non-variable frequency fan control model.

7. The method for controlling a non-variable frequency fan for a highway tunnel fire according to claim 6, characterized in that: Verifying the fan control parameters includes: Comparing the smoke spreading position with the fan position in the tunnel fixed parameters to determine the fan position intersecting with the smoke spreading position; According to the fan start position and the number of fan starts in the fan control parameters, determining whether there is a fan start in the fan position intersecting with the smoke spreading position; In response to the presence of a fan activation in a fan position intersecting the smoke spread location, the verification fails.

8. The method for controlling a non-variable frequency fan for a highway tunnel fire according to claim 6, characterized in that: Verifying the fan control parameters includes: If the verification is successful, the fan control parameter is used as an alternative control parameter, and the fan control parameter is continuously calculated by the non-variable frequency fan control model until the number of the alternative control parameters reaches a quantity threshold; The fan control parameter corresponding to the minimum control wind speed is selected from all the alternative control parameters.

Images