JP2703438B2 - Tunnel ventilation control device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tunnel ventilation control device for operating a plurality of ventilators installed in a tunnel to maintain the concentration of the contamination in the tunnel below an allowable value. It is possible to determine in a short time the operating combination of the ventilator necessary to maintain the value below the value, and to determine the appropriate operating combination following the change in the tunnel process characteristics over time even if the characteristics change over time. The present invention relates to a tunnel ventilation control device.

[0002]

2. Description of the Related Art Conventionally, in road tunnels, the exhaust gas from automobiles contaminates the inside of tunnels. Therefore, in long tunnels or tunnels with heavy traffic, mechanical ventilation is required to maintain the pollution concentration below an allowable value. Is being done.

On the other hand, many long tunnels have been constructed with the recent improvement of the road network, particularly the expressway network. In this long tunnel, the capacity of the installed ventilator is large and the number of installed ventilators is large. Therefore, if all the units are constantly operated, the required power becomes enormous. Therefore, in order to perform economical ventilation that requires less power, it is necessary to determine an appropriate operation combination so that only the necessary ventilator is operated according to the ventilation load.

Conventionally, as a method of determining the operation combination of the ventilator, for example, a method of storing several types of operation combination patterns corresponding to traffic levels in a minicomputer or a microcomputer in the form of a table, A method of determining a driving combination by mathematical programming based on a mathematical model of a process has been proposed.

[0005] However, the former table storage method has an advantage that a combination can be determined in a short time, but it is difficult to consider a plurality of factors of three or more variables, and it is not possible to take detailed measures. Therefore, an operation combination that does not fit the actual situation may be selected such that the ventilator is operated more than necessary, or conversely, the number of operated units is small and the pollution concentration is deteriorated.

On the other hand, when the latter mathematical programming method is used, it is necessary to solve the combination optimization problem. Therefore, when the number of ventilators is large, the number of combinations is large, and the calculation time is very long. There is a problem that it becomes longer.

In the mathematical programming method, a process model is used. However, in the case of a tunnel, it is difficult to create a highly accurate process model because the contamination concentration process is a distributed constant system and complicated.

Further, neither of the above methods can cope with aging of the tunnel process characteristics. That is, even if a good driving combination result is obtained at the beginning of operation, traffic conditions and process characteristics change over a long period of time after the operation starts, so that an appropriate result cannot be obtained.

[0009]

As described above, in the tunnel ventilation by the conventional method of determining the operation combination of the ventilator, the operation combination of the ventilator necessary to maintain the concentration of the contamination below the allowable value is determined. However, there is a problem that it takes a long time, and it is not possible to determine an appropriate operation combination following changes in the characteristics of the tunnel process over time.

[0010] It is an object of the present invention to be able to determine in a short time the operating combination of the ventilator required to maintain the pollution concentration below the permissible value, and to prevent the aging of the tunnel process characteristics. An object of the present invention is to provide an extremely reliable tunnel ventilation control device capable of executing an appropriate operation combination determination following a characteristic change.

[0011]

SUMMARY OF THE INVENTION In order to achieve the above object, the present invention is directed to a tunnel ventilation system for operating a plurality of ventilators installed in a tunnel to maintain the concentration of contamination in the tunnel below an allowable value. In the control device, a traffic volume measuring means for measuring a traffic volume passing through the tunnel, a traffic volume data storing device for storing a measured value from the traffic volume measuring device, and a data stored in the traffic volume data storing device. Based on the traffic volume prediction means for predicting the traffic volume passing through the tunnel, and the traffic volume predicted by the traffic volume prediction means, and a ventilator operation combination determination for determining a ventilation system operation combination pattern by a neural network Based on the means, the traffic volume predicted by the traffic volume prediction means, and the ventilator operation combination pattern determined by the ventilator operation combination determination means. Ventilation means for calculating the air volume of each ventilator necessary to maintain the contamination concentration in the ventilator below the allowable value, and ventilator operation result storage means for storing the operation results of the ventilator, Pollution concentration storage means for storing the measured values from the pollution concentration meter installed in the tunnel, traffic flow measurement data stored in the traffic flow data storage means, and ventilator operation results storage means Teaching data for learning the neural network used in the ventilator operation combination determining means based on the stored actual pattern of the ventilator operation combination and the contamination concentration data stored in the contamination concentration storage means And a neural network for learning the parameters of the neural network based on the teaching data created by the teaching data creating means. It is constituted by a click learning means. Here, an operation combination switching means for switching the operation combination of the ventilator based on the distributed ventilation volume calculated by the ventilation volume distribution means is added.

[0012]

Therefore, in the tunnel ventilation control device of the present invention, the traffic volume measured by the traffic volume measuring device is sent to the traffic volume data storage device and stored. Further, the traffic volume prediction means predicts an average traffic volume up to several tens of minutes based on data stored in the traffic volume data storage means. Further, in the ventilator operation combination determination means, the traffic volume predicted by the traffic volume prediction means is input, and the operation combination pattern of the ventilator is determined using the neural network.

On the other hand, in the ventilation amount distribution means, based on the traffic volume predicted by the traffic volume prediction means and the operation combination pattern determined by the ventilator operation combination determination means, the pollutant concentration is equal to or less than an allowable value and The air volume of each ventilator is calculated so that the required power is minimized. In addition, as needed, the operation combination switching means actually switches the operation combination of the ventilator based on the distribution result by the ventilation amount distribution means.

On the other hand, the ventilator operation result storage means stores the actual value of the ventilator operation pattern, and the pollution concentration storage means stores the measurement value from the pollution concentration meter installed in the tunnel. Then, in the teaching data creating means, the traffic flow measurement data stored in the traffic data storage means, the actual operation pattern of the ventilator stored in the ventilator operation result storage means, and the traffic concentration data stored in the pollution concentration storage means. Based on the pollutant concentration data, teaching data for learning the neural network used in the ventilator operation combination determination means is created. Further, the neural network learning means learns the parameters of the neural network by using the teaching data created by the teaching data creating means.

[0015]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below in detail with reference to the drawings. FIG. 2 is a diagram schematically illustrating a configuration of a road tunnel and a configuration of ventilation equipment to which the tunnel ventilation control device according to the present invention is applied.

In FIG. 2, a road tunnel 1 is a face-to-face traffic, and vehicles flow in both directions from the wellhead 1a to the wellhead 1b and from the wellhead 1b to the wellhead 1a. The ventilation wind is flowing from the wellhead 1a to the wellhead 1b.

On the other hand, as the ventilator, an electric dust collector 2,
3. The exhaust fan 5 and the blower 6 installed in the shaft 4, and the jet fan 13 installed in the ceiling of the road tunnel 1.
There is. The electric dust collectors 2 and 3 are devices for removing soot. The exhaust fan 5 is a device for discharging the polluted air in the road tunnel 1 to the outside of the road tunnel 1 through the exhaust-side duct of the shaft 4. Further, the blower 6 is a device for sending the air outside the road tunnel 1 into the road tunnel 1 through the blower-side duct of the shaft 4. Furthermore, the jet fan 13 is a device for flowing ventilation wind from the wellhead 1a side to the wellhead 1b side.

On the other hand, the traffic volume measuring device 7 measures the traffic volume traveling from the wellhead 1a to the wellhead 1b, and
Measures the traffic volume in the opposite direction. In addition,
Traffic volume is measured for large vehicles and small vehicles. The degree of visibility in the road tunnel 1 is determined by a fume transmittance meter (hereinafter, V
(Referred to as an I meter) 9 to 12. Then, the measured values of the traffic flow meter 7 and the VI meters 9 to 12 and the actual value of the operating airflow of the ventilator are all input to the ventilation control device 16.

Further, based on the input information, the ventilation control device 16 controls the electric precipitators 2, 3, and 3 so as to maintain the fume transmittance (hereinafter referred to as VI value) within a set range. It controls the air volume of the blower 5 and the blower 6 and the number of operating jet fans 13. The VI value is an index indicating the degree of visibility in the road tunnel 1, and is usually 1
It is represented by the transmittance of light within the 00m section. The smoke density is lower as the VI value is closer to 100%, and the smoke density is higher and the visibility is lower as the VI value is closer to 0%. In this embodiment, this VI value is controlled.

FIG. 1 is a functional block diagram showing a detailed configuration example of the ventilation control device 16 in FIG. FIG.
In FIG. 2, a plurality of traffic flow meters 7, 8, a barometer, a ventilator (electric dust collectors 2, 3, a blower 5, a blower 6), and VI meters 9 to 12 are respectively provided. 17,
The barometer 18, the ventilator 23, and the pollution concentration meter 32 are collectively expressed in boxes. In FIG. 1, the traffic data storage means 24 stores the measured value from the traffic
This is for saving as traffic data for the past month. The barometric pressure data storage means 25 stores the barometer 18
This is for storing the measured value from as atmospheric pressure data for one month.

Furthermore, the traffic volume prediction means 19 predicts the average traffic volume up to several tens of minutes passing through the road tunnel 1 for each of the large and small vehicles based on the data stored in the traffic data storage unit 24. Is what you do.

On the other hand, the ventilator operation combination determination means 20
Traffic volume and barometer 1 predicted by traffic volume prediction means 19
Input the measured value from 8 and use the neural network to set the operation combination pattern of the ventilator (combination of operation stop)
Is determined.

Further, the ventilation volume distribution means 21 calculates the traffic volume predicted by the traffic volume prediction means, the measured value from the barometer 18, and the ventilator operation combination pattern determined by the ventilator operation combination determination means 20. Based on the above, the air flow of each ventilator required to maintain the concentration of contamination in the road tunnel 1 below the allowable value and to minimize the required power of the ventilator is calculated.

Further, the operation combination switching means 22 determines whether the ventilator 2
The three driving combinations are actually switched. The ventilator 23 includes the electric precipitators 2 and 3, the exhaust fan 4, and the blower 5 shown in FIG. On the other hand, the ventilator operation result storage means 30 is for storing the actual value of the operation pattern of the ventilator 23 for one month. In addition, the pollution concentration data storage means 29 is for storing the measurement values from the pollution concentration meter installed in the road tunnel 1 for one month in the same manner.

Further, the teaching data creating means stores the traffic volume measurement data stored in the traffic volume data storage unit 24, the barometer data stored in the barometric pressure data storage unit 25, and the pollution concentration storage unit 29. Based on the stored pollution concentration data and the actual operation pattern of the ventilator 23 stored in the ventilator operation result storage means 30,
This is to create teaching data for learning the neural network used in the ventilator operation combination determination means 20.

Further, the neural network learning means 31 learns the parameters (weight coefficients) of the neural network based on the teaching data created by the teaching data creating means. Next, the operation of the tunnel ventilation control device of the present embodiment configured as described above will be described.

In FIG. 1, the traffic volume measuring device 17 detects vehicles passing through the tunnel for large vehicles and small vehicles, and sends the integrated value of the number of vehicles passing for 5 minutes to the traffic volume storage means 24. In this traffic volume storage means 24, the integrated value for 5 minutes is further integrated for one hour, and the past one month is stored as the hourly traffic volume [vehicle / h]. In addition, an average value for the past month is calculated in hourly units so as to be used as a basic pattern for traffic volume prediction. At this time, since the variation pattern of the traffic volume varies depending on the day of the week, the average value is calculated separately for weekdays, before holidays, holidays (Sunday and public holidays), and holidays. Next, the traffic volume prediction means 19 predicts a deviation from this average value in an hourly cycle using the following autoregressive model.

ΔNY _{k + 1} = A ₀ · ΔN _k + A ₁ · ΔN _k-1 + A ₂ · ΔN _k-2 (1) ΔN _k = N _k -NH _k (2) A _{0 to} A ₂ : coefficient ΔN _k : deviation from the average value of the actually measured traffic volume at the k-th time [vehicle / h] N _k : actual measured traffic volume at the k-th time [vehicle / h] NH _k : average traffic volume [vehicle / h] ΔNY _{k + 1} : Deviation forecast value one hour ahead [vehicle / h]

Therefore, the predicted traffic volume NY one hour ahead
_{k + 1} [unit / h] is calculated by the following equation. NY _{k + 1} = NY _{k + 1} + ΔNY _{k + 1} (3)

It should be noted that the prediction is made separately for a large car and a small car, going up and down. Also, the coefficient A _{0 of the} autoregressive model
To A _2, the sequential estimation is performed by the Kalman filter. The method of successively estimating the autoregressive model using the Kalman filter has already been described in Japanese Patent Application No. 63-58234, and will not be described here.

Next, the ventilator operation combination determination means 20
It is composed of a three-layer neural network as shown in FIG. The input data to the neural network are the up traffic volume x ₁ , the up traffic large vehicle mixing ratio x ₂ , the down traffic volume x ₃ , the down traffic large vehicle mixing ratio x _4, and the pressure difference between both pits x ₅ . The pressure difference between the two wells is the difference between the pressures measured by the barometers 14 and 15 shown in FIG.
However, these input data are normalized values as follows.

X ₁ = upward traffic volume predicted value / upstream traffic volume maximum value (4) x ₂ = upward large vehicle number predicted value / upstream traffic volume predicted value (5) x ₃ = downward traffic volume predicted value / downstream traffic maximum value ... (6) x ₄ = downlink large vehicle number predicted value / downlink traffic amount prediction value ... (7) x ₅ = (measured pressure difference - the minimum air pressure difference) / (maximum pressure difference - the minimum air pressure difference) ... (8)

Then, the ventilator operation combination determination means 20 includes an electric dust collector (two units) having a large equipment capacity and
The operation combination of the exhaust fan is determined. It is assumed that only one of the shaft blower and the exhaust fan is not operated. Therefore, there are eight types of operation combination patterns of the ventilator as shown in FIG. A jet fan having a small device capacity and capable of starting and stopping relatively easily is assumed to increase or decrease the number of operating units by feedback control, and the combination pattern determined by the ventilator operation combination determining means 20 includes a jet fan. Not included.
Since the feedback control is out of the scope of the present invention, the description thereof is omitted, but the method includes PID (proportional, integral, differential) control and fuzzy control.

The range of values taken by the output z _k (k = 1... 8) of the neural network shown in FIG.
In this embodiment, an operation pattern corresponding to the variable having the largest value (close to 1) is selected from the eight outputs.
The driving combination is determined. For example, larger values of z ₃ is the most, operation pattern number is 3, and becomes the operation of only transmitting and exhauster of the vertical shaft. The output z _k of the output layer of the neural network is calculated as follows. First, the output y _{j of the} hidden layer is calculated by the following equation. y _j = f _j (u _j ) (j = 1,..., n) (9)

[0035]

(Equation 1)

Here, y _j : output of the _hidden layer w _ji : weighting factor between the i-th neuron of the input layer and the j-th neuron of the hidden layer x _i : output of the i-th neuron of the input layer (the input and the output of the input layer are the same) f: a function representing the input / output characteristics of the neuron. Next, similarly to the case of the intermediate layer, the output z _{k of the} output layer is calculated as follows. z _k = f _k (u _k ) (k = 1,..., p) (11)

[0037]

(Equation 2)

Here, z _k : output of the output layer w _kj : weight coefficient between the k-th neuron of the intermediate layer and the k-th neuron of the output layer. As the function f, for example, a sigmoid function such as the following equation is used. f (u) = 1 / (1 + e ^-u ) (13)

The weighting factors in the neural network are learned by the neural network learning means 31, which will be described later, through actual operation. At first, it is necessary to set a value by some method. In this case, for example, the following method is conceivable.

First, the desired output T = (T ₁ ,..., T ₈ ) with respect to the input data X = (X ₁ ,..., X ₅ ) is calculated using the mathematical programming as described above. And
The weight coefficient w is obtained by back propagation using the obtained plural pairs of X and T as teaching data. If this weighting factor has been obtained, the output of the neural network simply calculates equations (9) to (12), and the driving combination can be determined in a very short time.

Next, a back propagation method for learning a weight coefficient will be described. Here, the error function in the output layer is defined as in equation (14).

[0042]

(Equation 3)

Here, Z _k : output of the k-th neuron in the output layer T _k : teaching data (desired value) for the k-th neuron in the output layer. Backpropagation is a method of correcting the weight coefficient so that the error function approaches the minimum value. The correction values Δw _kj and Δw _ji of the weight coefficient are calculated by the following equations.

[0044]

(Equation 4)

In the above equation, ε is a parameter that determines the magnitude of one correction. When the partial differential terms of Expressions (15) and (16) are expanded and arranged, the amount of correction of the weight coefficient is as follows.

Δw _kj = −ε · δ _k · y _j (17) Δw _ji = −ε · δ _j · x _i (18)

[0047]

(Equation 5) Further, in order to suppress the oscillation at the time of learning and accelerate the convergence of the learning, a method using the following expression instead of the expressions (17) and (18) is also effective.

[0048] _{Δw kj (t) = -ε ·} δ k · y j + α · Δw kj (t-1) ... (21) Δw ji (t) = -ε · δ j · x i + α · Δw ji (t −1)… (22) (0 <α <1)

Here, α is a parameter for stabilizing learning, and t represents the number of times of learning. The details of the neural network and the back propagation are described in, for example, the following documents.
"Edited by Kaoru Nakano:" Basics of Neuron Computer ", Corona Publishing (1980)"

Next, in the ventilation volume distribution means 21, the traffic volume predicted by the traffic volume prediction device 19, the measured value from the barometer 18, and the operation of the ventilator determined by the ventilator operation combination determination device 20. The air volume of the operating ventilator is calculated based on the stop combination. As the ventilators of the present embodiment, there are two electric precipitators and a blower and a blower of a shaft. When calculating the airflow of the ventilator, a nonlinear programming method is used to satisfy the constraints such as the upper limit of the concentration of pollution and the upper and lower limits of the airflow of the ventilator and the minimum power requirement of the ventilator. Is done. Here, the nonlinear programming problem is formulated as follows.

(1) Variables X ₁ : Air volume of the first dust collector [m ₃ / sec] X ₂ : Air volume of the second dust collector [m ₃ / sec] X ₃ : Air volume of the exhaust air [m ₃ / sec] X ₄ : Air volume of the air blower _{[m 3 / sec] X 5} : roadway air volume wellhead 1a side of the shafts _{[m 3 / sec] X 6} : wellhead 1b side of the roadway airflow than shafts _{[m 3 / sec] X 7} : number of operating a jet fan [units ]

(2) Objective function F = W ₁ (X ₁ ) + W ₂ (X ₂ ) + W ₃ (X ₃ ) + W ₄ (X ₄ ) + W ₅ (X ₇ ) → Minimization (23) W ₁ : Function for finding the required power of the first dust collector [kw] W ₂ : Function for finding the required power of the second dust collector [kw] W ₃ : Function for finding the required power of the exhaust fan [kw] w ₄ : Finding the required power of the blower function [kw] w _5: function for obtaining the required power of the jet fan [kw]

(3) Inequalities constraint conditions Upper and lower limit of dust collector air volume Q _C1MIN ≦ X ₁ ≦ Q _C1MAX (24) Q _C2MIN ≦ X ₂ ≦ Q _C2MAX (25) Q _C1MIN : Minimum air volume of the first dust collector [m ³ / sec] _QC2MIN : Minimum air volume of the second dust collector [m ³ / sec] _QC1MAX : Maximum air volume of the first dust collector [m ³ / sec] Q _C2MAX: second collector maximum air volume [m ³ / sec]

The upper and lower limits of the air volume of the blower / discharger Q _EMIN ≦ X ₃ ≦ Q _EMAX (26) Q _BMIN ≦ X ₄ ≦ Q _BMAX (27) Q _EMIN : Minimum air volume of the blower [m ³ / sec] Q _BMIN : Minimum air volume of blower [m ³ / sec] Q _EMAX : Maximum exhaust air volume [m ³ / sec] Q _BMAX : Maximum air volume of blower [m ³ / sec]

Restrictions on the Number of Jet Fans Operating MJF _MIN = X ₇ ≦ NJF _MAX (28) MJF _MIN : Minimum Number of Jet Fans Operating [Unit] NJF _MAX : Maximum Number of Jet Fans Operating [Unit]

Other restrictions on air volume X ₁ ≦ X ₅ (29) X ₂ ≦ X ₆ (30) X ₃ ≦ X ₅ (31) X ₄ ≦ X ₆ (32)

[0057] Soot concentration constraint _{C VI1 (N L, N S} , X 5) ≦ C VIMAX ... (33) C VI2 (N L, N S, C VI1, X 1, X 5) ≦ C VIMAX ... (34) _{C VI3 (N L, N S} , C VI2, X 3, X 4, X 6) ≦ C VIMAX ... (35) C VI4 (N L, N S, C VI3, X 2, X 6) ≦ C VIMAX ... (36) C _VIi: soot concentration of the iVI meter installation point [-] N _L: large car number [units / h] N ^S : Number of small vehicles [vehicle / h] C _VIMAX : upper limit of soot concentration [-] Here, C _{VI1 to} C _VI4 are functions. Further, there is a relationship as shown below between the VI value and the soot concentration.

C _VI = (− 1/100) · log ₁₀ (VI / 100) (37) C _VI : soot concentration [−] VI: VI value [%] Therefore, VI lower limit of VI in the expression (37) is obtained. By substituting the value [%], the soot concentration upper limit value C _VIMAX is calculated.

[0059] (4) equality constraints pressure balance _{ΔP t1 + ΔP t2 + ΔP m} + ΔP r1 + ΔP r2 + ΔP c1 + ΔP c2 + ΔP B + ΔP JF = 0 ... (38) ΔP t1 = f1 (N L, N S, X 5) _{... (39-1) ΔP t2 = f2} (N L, N S, X 6) ... (39-2) ΔP r1 = f3 (X 5) ... (39-3) ΔP r2 = f4 (X 6) ... ( 39-4) ΔP _c1 = f5 (X ₁ , X ₅ ) (39-5) ΔP _c2 = f6 (X ₂ , X ₆ ) (39-6) ΔP _B = f7 (X ₄ , X ₆ ) _{(39-7) ΔP JF = f8 (} X 7) ... (39-8) ΔP t1: traffic ventilation force wellhead 1a side of the shafts [mmAq] ΔP _t2: traffic ventilation force wellhead 1b side of the shafts [mmAq] [Delta] P _m: both wellhead between pressure difference [mmAq] ΔP _r1: carriageway resistance wellhead 1a side of the shafts [mmAq] ΔP _r2: carriageway resistance wellhead 1b side of the shafts [mmAq] [Delta] P _c1 Boosting power of the first dust collector [mmAq] ΔP _c2: boosting power of the second collector [mmAq] ΔP _B: boosting power of the blower [mmAq] ΔP _JF: boosting power of the jet fan [mmAq] Here, f1 to f8 function It is.

Air volume balance X ₅ −X ₃ + X ₄ = X ₆ (40) In the nonlinear programming, the variable X ₁ that satisfies the constraints (24) to (40) and minimizes the objective function (23) ~ X
₇ is required.

The nonlinear programming is a generally known technique, and is described, for example, in the following literature. "Kono, Yamashita:" Nonlinear Programming ", Nikkagiren (S53)"

The formulas for calculating the soot concentration in (33) to (36) and the formulas for calculating the pressure in (39-1) to (39-8) are described in the following documents. "Japan Road Association:" Road Tunnel Technical Standards (Ventilation) / Commentary ", Japan Road Association (S60)"

Next, the operation combination switching means 22 operates each of the ventilators in accordance with the calculation results of the ventilation amount distribution means 21 (the operation air amount of the electric dust collector, the operation air amount of the exhaust fan, and the operation air amount of the blower). The air flow command is output, and the operation combination is switched.

The operation results of the ventilator are stored in the ventilator result data storage means 30. This stored data is the operating air volume of each ventilator. Further, the measured value from the contamination concentration meter is stored in the contamination concentration data storage unit 29. In the case of the present embodiment, the measurement values from the VI meters 9 to 12 are stored.

On the other hand, if the process characteristics of the tunnel change over time, if the weighting factor of the neural network used in the ventilator operation combination determination means 20 is fixed, the operation combination determined May not match the actual situation. Therefore, teaching data for learning the weight coefficient is created in the teaching data creating unit 27.

That is, in the teaching data creating means 27,
Based on data stored in the traffic volume storage unit 24, the atmospheric pressure data storage unit 25, the pollution concentration storage unit 29, and the ventilator operation result storage unit 30, teaching data for learning the weighting factor of the neural network is created. Is done. The teaching data, the input variables shown in FIG. 3 (x ₁ ~x ₅₎
Plurality created output variables (z ₁ ~z ₈₎ as a pair and. Teaching data means that when given input data,
It shows the desired output value. The output variables z _{1 to} z ₈ are variables for selecting an operation pattern, and an operation pattern corresponding to a variable whose value is closest to 1 is selected. Therefore, the output variable in the teaching data is
Only one has a value of 1 and all others have a value of 0. Next, an example of creating teaching data will be described with reference to FIG.

The upper graph in FIG. 5 shows the change in the VI value, which is data stored in the contamination concentration storage means 29. The lower graph shows the change in the ventilator operation pattern.
3 is a graph created from data stored in the.
As the operation result data of the ventilator, the operation air volume of each ventilator is stored. Since the operation of the ventilator is known from the operation air volume data, it is converted into the pattern number of the operation combination, and is converted into the operation combination shown in FIG. A graph like the one below is created.

Normally, the ventilator is operated according to the operation pattern determined by the above-described ventilator operation combination determination means 20. However, when the VI value is abnormally reduced or in the case of excessive ventilation, emergency value control or The operation combination of the ventilator is changed by manual control. In Figure 5, the operation pattern by manual control at time t ₂ has been changed from 5 8.

As the teaching data, it is desirable that there is no abnormal decrease in the VI value or excessive ventilation, and the VI value falls within a certain range. As shown in the upper graph of FIG.
Here, a lower limit value VI _MIN and an upper limit value VI _MAX are set as the range of the VI value. Then, an operation pattern in a time zone in which the frequency is frequently out of the range of the upper and lower limits is not adopted as the teaching data. In this example, the operation pattern of the time zone t ₁ ~t ₂ is excluded from the teaching data.
To the subject of the teaching data is the time period of 0 to t ₁ and t ₂ ~t _3. For the target time zone, the average value of the data stored in the traffic volume data storage unit 24 and the average value of the data stored in the atmospheric pressure data storage unit 25 are calculated, and the input variables of the teaching data are calculated. (X _{1 to} x
₆ ) is created. The teaching data for the output variables is as follows. 0~t ₁ time _{zone: z 3 = 1, z k} = 0 (k ≠ 3) t 2 ~t 3 of the time _{zone: z 8 = 1, z k} = 0 (k ≠ 8)

Further, in this embodiment, the input / output device 26 can be used to manually correct or add teaching data. Then, the created teaching data is stored in the teaching data storage unit 28. When creating the teaching data in the teaching data creating means 27, for example, weather data may also be input from the input / output device 26 in addition to the above-described data, and may be used as one element of the teaching data creation.

Next, the neural network learning means 3
In step 1, learning of the weight coefficient of the neural network used in the ventilator operation combination determination means 20 is performed based on the teaching data stored in the teaching data storage means 28. In this case, the learning method is back propagation. This back propagation method is as described in the description of the ventilator operation combination determination means 20. And this neural network learning means 31
Is used in the neural network of the ventilator combination determination means 20.

As described above, in the tunnel ventilation control device of this embodiment, the traffic volume and the data of the barometer are input, and the ventilator (electric dust collector and blower / discharger) is used by using a three-layer neural network. In order to determine the driving combination, and to create teaching data from online operation data, and to automatically learn the weight coefficient of the neural network by backpropagation, the contamination concentration is set to an allowable value or less. It is possible to determine the operation combination of the ventilator necessary to maintain it in a very short time, and even if the secular change of the tunnel process characteristics is large, execute the appropriate combination determination following the characteristic change Is possible.

In the above embodiment, the case where the present invention is applied to a face-to-face traffic tunnel has been described. On the other hand, in the case of a traffic tunnel, the traffic volume input to the neural network and the mixing ratio of large vehicles are reduced by one. What is necessary is just to make it data only in the direction (up or down).

In the above embodiment, the case where a barometer is installed has been described. However, in reality, there are many tunnels in which barometers are not installed at both pits. What is necessary is just to delete the measured value of the barometer from the input data of. That is, in the neural network shown in FIG. 3, to remove the pressure difference x ₅ between the wellhead which is an input variable, the number of neurons in the input layer 4
You only have to reduce it to individual pieces. Further, in the above embodiment, the case where the operation combination switching means 22 is provided has been described, but this is not essential to the present invention.

[0075]

As described above, according to the present invention, data on the traffic volume in a tunnel is input, the operation combination of the ventilator is determined using a neural network, and the data is taught from online operation data. Since the data was created and the weight coefficient of the neural network was automatically learned by back propagation,
The operating combination of the ventilator required to maintain the pollutant concentration below the allowable value can be determined in a short time, and even if the characteristics of the tunnel process change over time, appropriate operation can be performed following the change in the characteristics. An extremely reliable tunnel ventilation control device capable of executing the combination determination can be provided.

[Brief description of the drawings]

FIG. 1 is a functional block diagram showing an embodiment of a ventilation control device according to the present invention.

FIG. 2 is a schematic configuration diagram showing an example of a road tunnel in the embodiment.

FIG. 3 is a diagram showing an example of a neural network in the embodiment.

FIG. 4 is a diagram showing an example of a pattern of a ventilator operation combination in the embodiment.

FIG. 5 is a diagram showing an example of operation data in the embodiment.

[Explanation of symbols]

1: Road tunnel, 2, 3: Electric dust collector, 4: Vertical shaft, 5
... exhaust fan, 6 ... blower, 7, 8 ... traffic meter, 9-1
2: soot transmittance meter (VI meter), 13: jet fan,
14, 15: barometer, 16: ventilation control device, 19: traffic volume prediction means, 20: ventilator operation combination determination means, 21 ...
Ventilation volume distribution means, 22 ... driving combination switching means, 24 ... traffic data storage means, 25 ... atmospheric pressure data storage means, 26
... input / output device, 27 ... teaching data creation means, 28 ... teaching data storage means, 29 ... pollution concentration storage means, 30 ... ventilator operation result storage means, 31 ... neural network learning means.

Claims (2)

(57) [Claims] 1. A tunnel ventilation control device for operating a plurality of ventilators installed in a tunnel to maintain the concentration of contamination in the tunnel below an allowable value, wherein the traffic measuring the traffic passing through the tunnel is controlled. Traffic measuring means, a traffic data storing means for storing the measured value from the traffic measuring means, and a traffic passing through the tunnel based on the data stored in the traffic data storing means. Traffic volume prediction means to predict, input the traffic volume predicted by the traffic volume prediction means,
A ventilator operation combination determining means for determining an operation combination pattern of the ventilator by a neural network; a traffic volume predicted by the traffic volume prediction means; and a ventilator operation combination determined by the ventilator operation combination determination means. Based on the pattern, ventilation volume distribution means for calculating the air volume of each ventilator required to maintain the concentration of contamination in the tunnel below an allowable value, and for storing the operation results of the ventilator Ventilator operation result storage means, pollution concentration storage means for storing a measured value from a pollution concentration meter installed in the tunnel, traffic volume measurement stored in the traffic volume data storage unit Data, the actual pattern of the ventilator operation combination stored in the ventilator operation result storage means, and the contamination concentration data stored in the contamination concentration storage means. Based on
Teaching data creating means for creating teaching data for learning the neural network used in the ventilator operation combination determining means; and a parameter of the neural network based on the teaching data created by the teaching data creating means. And a neural network learning means for learning. 2. The tunnel ventilation control device according to claim 1, further comprising an operation combination switching unit that switches an operation combination of the ventilator based on the distributed ventilation calculated by the ventilation distribution unit. A tunnel ventilation control device, characterized in that: