Tunnel monitoring system

Abstract

To obtain a tunnel monitoring system that can accurately identify conditions within a tunnel on the basis of acoustic data acquired inside the tunnel serving as a monitoring area.SOLUTION: A tunnel monitoring system comprising a plurality of sound collecting devices that are arranged in a line in both end directions of a tunnel in the tunnel that is a monitoring area and that respectively collect acoustic data generated in the tunnel, and an acoustic data processing unit that calculates sound pressure levels for each frequency band by performing frequency analysis on the acoustic data collected by each of the plurality of sound collecting devices, wherein the acoustic data processing unit identifies a type of sound generated in the tunnel and a generation position thereof based on results of the frequency analysis and an arrangement of the plurality of sound collecting devices.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a tunnel monitoring system that creates and displays display data for grasping the situation inside a tunnel, which is a monitoring area, based on acoustic data collected inside the tunnel. [Background technology]

[0002] There is an abnormal sound monitoring system that can accurately and easily detect abnormal conditions inside a building based on the results of collecting sounds made inside the building (see, for example, Patent Document 1). The abnormal sound monitoring system disclosed in Patent Document 1 extracts abnormal sounds from sound information collected inside the building, estimates the sound source position of the sound information, and calculates a danger value that indicates the degree of danger of the abnormal sound based on the estimation result.

[0003] As a result, the abnormal sound monitoring system disclosed in Patent Document 1 can calculate the danger value of an abnormal sound with high accuracy, taking into account various factors, based on information such as the type of abnormal sound and the location of the sound source. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Publication No. 2020-159966 Summary of the Invention [Problem to be solved by the invention]

[0005] However, the conventional techniques have the following problems. In Patent Document 1, it is necessary to extract abnormal sounds with high accuracy, but the accuracy of extracting abnormal sounds deteriorates depending on the monitoring area due to the influence of reflections, etc. In particular, when the monitoring area is a tunnel, the influence of sound reflections inside the tunnel makes it difficult to identify abnormal sounds and accurately detect the location of the sound source.

[0006] Furthermore, Patent Document 1 also displays the location of the abnormal sound and its movement. However, when a tunnel is used as the monitoring area, the detection accuracy of the sound source location itself deteriorates, and the reliability of the displayed content also deteriorates.

[0007] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a tunnel monitoring system that can identify the condition inside a tunnel with high accuracy based on acoustic data collected within the tunnel, which is the monitored area. [Means for solving the problem]

[0008] The tunnel monitoring system according to the present disclosure comprises a plurality of sound collection devices arranged in a row on both ends of the tunnel, which is the monitoring area, and which each collect sound data generated within the tunnel, and an acoustic data processing unit that calculates the sound pressure level for each band by performing frequency analysis on the sound data collected by each of the plurality of sound collection devices, and the acoustic data processing unit identifies the type and location of sound generated within the tunnel based on the results of the frequency analysis and the arrangement of the plurality of sound collection devices. [Effects of the Invention]

[0009] According to the present disclosure, it is possible to obtain a tunnel monitoring system that can identify the condition inside a tunnel with high accuracy based on acoustic data collected inside the tunnel, which is the monitoring area. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is an explanatory diagram illustrating an overall configuration of a tunnel monitoring system according to a first embodiment of the present disclosure. [Figure 2] FIG. 2 is an explanatory diagram showing a spectrogram that is a result of frequency analysis of acoustic data related to human vocalizations in the first embodiment of the present disclosure. [Figure 3]10 is a first display example in which display data generated based on a frequency analysis result is displayed by the acoustic data processing unit according to the first embodiment of the present disclosure. [Figure 4] 10 is a second display example in which display data generated based on a frequency analysis result by the acoustic data processing unit according to the first embodiment of the present disclosure is displayed. [Figure 5] FIG. 10 is an explanatory diagram showing a specific example of distinguishing between the sound of a person speaking and the sound of water discharge using display data generated based on a frequency analysis result by the acoustic data processing unit according to the first embodiment of the present disclosure. [Figure 6] 10 is an explanatory diagram showing a specific example of distinguishing between human vocalizations and warning sounds using display data generated based on frequency analysis results by the acoustic data processing unit according to the first embodiment of the present disclosure. FIG. [Figure 7] FIG. 10 is an explanatory diagram showing a specific example of distinguishing between human vocalization sounds and vehicle sounds using display data generated based on frequency analysis results by the acoustic data processing unit according to the first embodiment of the present disclosure. [Figure 8] 10 is an explanatory diagram illustrating a case where the attenuation direction of a vocal sound is identified from acoustic data collected by a plurality of microphones in the acoustic data processing unit according to the first embodiment of the present disclosure. FIG. [Figure 9] 4 is a flowchart showing a series of processes related to the creation of display data executed in an acoustic data processing unit in the tunnel monitoring system according to the first embodiment of the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, preferred embodiments of the tunnel monitoring system of the present disclosure will be described with reference to the drawings. The tunnel monitoring system of the present disclosure has a technical feature in that it calculates the sound pressure level for each band based on acoustic data collected from multiple sound collection devices within the tunnel, which is the monitoring area, and identifies and displays the sound pressure level for each band at each position of the multiple sound collection devices, allowing an operator who visually views the identified display results to grasp the condition inside the tunnel with high accuracy.

[0012] Embodiment 1 1 is an explanatory diagram showing the overall configuration of a tunnel monitoring system according to a first embodiment of the present disclosure. The tunnel monitoring system according to the first embodiment is configured to include a plurality of microphones 10 and an acoustic data processing unit 20.

[0013] The multiple microphones 10 correspond to sound collection devices that are respectively placed at different positions between the entrance and exit of a tunnel, which is the monitoring area. In Fig. 1, the multiple microphones 10 are configured as N microphones 10(1) to 10(N), and are placed in a line toward both ends of the tunnel.

[0014] Unlike a smoke detector, the microphone 10 does not need to be installed on the ceiling of the tunnel, allowing for flexibility in determining where to install it. In particular, for the purpose of collecting acoustic data for determining whether or not someone is inside the tunnel based on vocalizations, the microphone can be installed at a height suitable for collecting vocalizations. Specific examples of suitable heights include a range of 1.2 m to 1.7 m from the bottom of the tunnel where people walk.

[0015] The acoustic data processing unit 20 calculates the sound pressure level for each band by performing frequency analysis on the acoustic data collected by each of the multiple microphones 10. Furthermore, the acoustic data processing unit 20 creates display data that makes it possible to identify the sound pressure level for each band at the position of each of the multiple microphones 10, and displays the display data on the display unit. Here, the acoustic data processing unit 20 can be installed, for example, in a central control room that centrally monitors tunnels.

[0016] The acoustic data processing unit 20 can update the display data sequentially by using the average sound pressure level calculated at predetermined intervals as the sound pressure level and creating display data at each interval. For example, the acoustic data processing unit 20 can suppress the effects of noise by using 1 second as the predetermined interval, calculating the sound pressure level as the average value over 1 second, and updating the display data.

[0017] The tunnel monitoring system according to the first embodiment is characterized in that it generates display data based on acoustic data, thereby helping an operator who visually views the display data to infer the situation inside the tunnel, such as the sounds of people speaking, warning sounds, idling vehicles, and water discharge sounds.

[0018] In particular, it can be used to provide information to visually determine whether there are people in the tunnel who have been unable to escape in the event of an emergency such as a fire inside the tunnel. Therefore, a specific method for identifying and displaying the various sounds that occur inside the tunnel will be explained below using diagrams.

[0019] 2 is an explanatory diagram showing a spectrogram that is a result of frequency analysis of acoustic data related to human vocalizations in embodiment 1 of the present disclosure. In Fig. 2, the vertical axis represents frequency and the horizontal axis represents time, and time-series data related to the spectrogram of the acoustic data is displayed, with sound pressure levels being displayed in different colors.

[0020] The frequency band of human voices is large, between 100 and 2000 Hz. On the other hand, the sound of water being sprayed during firefighting activities, for example, is louder in the low frequencies and attenuates as the frequencies increase, depending on the amount of water being sprayed. Therefore, by visualizing these differences resulting from frequency analysis, it is possible to distinguish between speech sounds and water spray sounds.

[0021] 3 is a first display example in which display data generated based on frequency analysis results by the acoustic data processing unit 20 according to the first embodiment of the present disclosure is displayed. In the first display example, the vertical axis represents frequency, the horizontal axis represents locations corresponding to the installation positions of the multiple microphones 10, and the frequency analysis results of the acoustic data collected at each installation location are displayed as sound pressure levels for each band.

[0022] In the first display example shown in Fig. 3, sound pressure levels are displayed in different colors according to their magnitude. That is, in the first display example shown in Fig. 3, three-dimensional information consisting of three elements, frequency, position, and sound pressure level, is expressed by using frequency as the vertical axis, position as the horizontal axis, and sound pressure level as a different color.

[0023] Furthermore, Figure 3 shows the display data obtained as a result of generating a human vocalization sound at a position corresponding to location 4 and a water-spraying sound at a position corresponding to location 9, and then frequency-analyzing the acoustic data collected by multiple microphones 10.

[0024] As is clear from Figure 3, the highest sound pressure level for vocalization sounds was obtained at location 4 in the 100 to 2000 Hz band, and the highest sound pressure level for water discharge sounds was obtained at location 9 in the low frequency band of 0 to 10 Hz.

[0025] The acoustic data processing unit 20 can identify the location where the highest sound pressure level is obtained in the 100 to 2000 Hz band as the human detection location, and in the first display example shown in Figure 3, the words "human detection location" are displayed at location 4 along with a circle.

[0026] Therefore, by visually checking the first display data shown in FIG. 3, the operator in the central control room can determine that there is a high possibility that a person is present at location 4 in the tunnel.

[0027] 4 is a second display example in which display data generated based on the frequency analysis results by the acoustic data processing unit 20 according to the first embodiment of the present disclosure is displayed. In the second display example, the vertical axis represents sound pressure level, the horizontal axis represents locations corresponding to the installation positions of the multiple microphones 10, and the frequency analysis results of the acoustic data collected at each installation location are displayed as the highest sound pressure level for each location.

[0028] In the second display example shown in Fig. 4, the frequency centroids for each of the highest sound pressure levels are displayed in different colors according to their size. That is, in the second display example shown in Fig. 4, three-dimensional information consisting of three elements, sound pressure level, position, and frequency, is expressed by using sound pressure level as the vertical axis, position as the horizontal axis, and frequency centroids as color-coded displays.

[0029] Furthermore, in Figure 4, similar to Figure 3 above, the sound of a person speaking is generated at a position corresponding to location 4, and the sound of water being sprayed is generated at a position corresponding to location 9, and the display data obtained as a result of frequency analysis of the acoustic data collected by multiple microphones 10 at that time is shown.

[0030] The acoustic data picked up by the microphone 10 closer to the sound source has a higher sound pressure level. Therefore, as is clear from Figure 4, for the vocalization sound, the highest sound pressure level with a frequency center of gravity in the 100 to 2000 Hz band was obtained near the microphone 10 (5), and for the water discharge sound, the highest sound pressure level with a frequency center of gravity in the 0 to 10 Hz band was obtained near the microphone 10 (10).

[0031] Therefore, by visually checking the second display data shown in Figure 4, the operator in the central control room can determine that there is a high possibility that a person is near the installation location of microphone 10(5) in the tunnel.

[0032] Whether the first or second display example is adopted, it becomes possible to distinguish the sound pressure level for each band at the position of each of the multiple microphones 10, which is effective in detecting people inside a tunnel.

[0033] Next, using Figures 5 to 7, we will explain how the occurrence of each of the water discharge sound, warning sound, and vehicle sound in idling state can be distinguished from the occurrence of human vocalization sound by comparing the frequency analysis results. Note that Figures 5 to 7 will be explained assuming that the first display example shown in Figure 3 is used.

[0034] FIG. 5 is an explanatory diagram showing a specific example of distinguishing between the sound of a person speaking and the sound of water discharge using display data generated based on the frequency analysis results by the acoustic data processing unit 20 according to the first embodiment of the present disclosure.

[0035] In addition, Fig. 5 shows the display data obtained as a result of frequency analysis of the acoustic data collected by multiple microphones 10 when a human voice was generated at a position corresponding to location 4 and a water discharge sound was generated at a position corresponding to location 9. The upper part of Fig. 5 also shows a schematic representation of the positions of the people in the tunnel and the water discharge position.

[0036] A powerful water jet, such as that used to extinguish a fire, produces a sound pressure level that peaks in the low frequency range of 10 Hz or less. On the other hand, a voice produces a sound pressure level that peaks in the 100-2000 Hz band. Therefore, an operator visually checking Figure 5 can distinguish this from the sound of a water jet and determine that there is a high possibility that someone is present at location 4 in the tunnel.

[0037] FIG. 6 is an explanatory diagram showing a specific example of distinguishing between human vocalizations and warning sounds using display data generated based on frequency analysis results by the acoustic data processing unit 20 according to the first embodiment of the present disclosure.

[0038] 6 shows the display data obtained as a result of frequency analysis of the acoustic data collected by multiple microphones 10 when a human voice is generated at a position corresponding to location 4 and a warning sound is generated at a position corresponding to location 9. The upper part of FIG. 6 also shows a schematic representation of the positions of the people in the tunnel and the positions where the warning sound is played.

[0039] A warning sound, such as a siren that sounds when a fire breaks out, varies continuously from low to high frequencies over the 300 Hz to 4 kHz band, creating a sweeping sound with a frequency that changes over time. On the other hand, a spoken sound has a sound pressure level that peaks in the 100 to 2000 Hz band. Therefore, an operator visually viewing Figure 6 can distinguish this from a warning sound and determine that there is a high possibility that someone is present at location 4 in the tunnel.

[0040] FIG. 7 is an explanatory diagram showing a specific example of distinguishing between human vocalization sounds and vehicle sounds using display data generated based on frequency analysis results by the acoustic data processing unit 20 according to the first embodiment of the present disclosure.

[0041] 7 shows the display data obtained as a result of frequency analysis of the acoustic data collected by multiple microphones 10 when a human voice is generated at a position corresponding to location 4 and an idling vehicle sound is generated at a position corresponding to location 9. The upper part of FIG. 7 also shows a schematic representation of the positions of people and vehicles in the tunnel.

[0042] The sound of an idling vehicle peaks around 25 Hz. On the other hand, the sound pressure level of a spoken voice peaks in the 100 to 2000 Hz band. Therefore, an operator visually checking Figure 7 can distinguish this from the sound of a vehicle and determine that there is a high possibility that a person is present at location 4 in the tunnel.

[0043] 5 to 7 illustrate the case of identifying the position of a person in a tunnel. However, in the tunnel monitoring system according to the present disclosure, multiple microphones 10 are arranged in a line toward both ends of the tunnel, and sound data can be collected at different locations. Therefore, from the difference in sound pressure levels at each location, it is possible to estimate the direction in which the vocal sound is gradually attenuating.

[0044] 8 is an explanatory diagram of a case where the attenuation direction of a vocal sound is identified from acoustic data collected by a plurality of microphones in the acoustic data processing unit 20 according to the first embodiment of the present disclosure. In FIG. 8, 12 microphones 10(1) to 10(12) are arranged in a tunnel toward both ends of the tunnel.

[0045] The acoustic data processing unit 20 calculates the average sound pressure level in the frequency band of the vocal sound by performing frequency analysis on the acoustic data obtained by each of the microphones 10(1) to 10(12). In the example of Fig. 8, the frequency band of the vocal sound is set to 100 Hz to 2000 Hz, and the average sound pressure levels corresponding to each of the microphones 10(1) to 10(12) are calculated as 0 dB, 6 dB, 10 dB, 15 dB, 17 dB, 20 dB, 10 dB, 5 dB, 3 dB, 0 dB, 0 dB, and 0 dB.

[0046] Next, the acoustic data processing unit 20 identifies the microphone 10 that picked up acoustic data with the maximum value among the 12 average sound pressure levels calculated for each of the microphones 10(1) to 10(12) as the first sound collection device. In the example of Fig. 8, the maximum value of the average sound pressure level in the frequency band of the vocal sound is 20 dB, and the microphone 10(6) that picked up the maximum value of 20 dB is identified as the first sound collection device.

[0047] Next, the sound data processing unit 20 identifies, as the second sound collection device, the microphone 10 that collected sound data having an average sound pressure level within a preset sound pressure attenuation amount from the maximum value. In the example of Fig. 8, when the maximum value is 20 dB and the preset sound pressure attenuation amount is 6 dB, the microphones 10(4) and 10(5) that collected sound data having an average sound pressure level of 14 dB or more are identified as the second sound collection device.

[0048] Next, the acoustic data processing unit 20 creates display data that further distinguishably displays the range from the position of the microphone 10(6) identified as the first sound collection device to the positions of the microphones 10(4) and 10(5) identified as the second sound collection devices. In the example of Fig. 8, an arrow for distinguishably displaying the range from the microphone 10(4) to the microphone 10(6), with the microphone 10(6) as the start point and the microphone 10(4) as the end point, is shown as part of the display data.

[0049] Therefore, an operator who further visually recognizes the identification display using the arrow as shown in Figure 8 can infer that a person is present at the position of microphone 10(6) and is speaking in the direction of microphone 10(4). Furthermore, by visually recognizing the successively updated display data including such identification display using the arrow in chronological order, the operator can infer in which direction the person is moving.

[0050] In the specific example described above, the operator visually checked the display data and determined whether or not the data contained characteristics of a vocalization. However, if necessary, it is also possible to quantitatively determine whether or not the audio data contains a vocalization by performing voice recognition processing on the audio data.

[0051] In this case, the acoustic data processing unit 20 determines whether or not there is acoustic data containing human vocalizations by executing pre-trained voice recognition processing on the acoustic data collected by each of the multiple microphones 10. If the acoustic data processing unit 20 determines that there is acoustic data containing vocalizations, it can identify the position of the microphone 10 that collected the acoustic data containing the vocalizations as the position of the person in the tunnel and display the identification result on the display unit.

[0052] The position identified by executing the voice recognition process can be displayed as a "person detection position" in a manner similar to that shown in FIG.

[0053] Next, a series of processes executed in the acoustic data processing unit 20 according to the first embodiment will be described using a flowchart. Fig. 9 is a flowchart showing a series of processes related to the creation of display data executed in the acoustic data processing unit 20 in the tunnel monitoring system according to the first embodiment of the present disclosure.

[0054] First, in step S901, the acoustic data processing unit 20 performs a process of collecting acoustic data generated at each location in the tunnel, which is the monitoring area, via a plurality of microphones 10 installed in the tunnel.

[0055] Next, in step S902, the sound data processing unit 20 performs frequency analysis on the sound data collected by each of the plurality of microphones 10 to calculate the sound pressure level for each band at the installation position of each microphone 10.

[0056] Next, in step S903, the acoustic data processing unit 20 creates display data based on the calculation results regarding the sound pressure level for each band. Specifically, the acoustic data processing unit 20 can create the first display example shown in Figures 3 and 5 to 7 or the second display example shown in Figure 4 as display data, and can also include in the display data an identification display indicating the range of vocal sounds shown in Figure 8.

[0057] Next, in step S904, the acoustic data processing unit 20 determines whether or not a setting is made to perform processing for identifying a person's position based on voice recognition processing as an optional function. If the setting is to perform voice recognition processing, the acoustic data processing unit 20 executes the processing from step S905 onward, and if the setting is not to perform voice recognition processing, the acoustic data processing unit 20 executes the processing of step S907.

[0058] If the process proceeds to step S905, the acoustic data processing unit 20 performs pre-trained voice recognition processing on the acoustic data collected by each of the multiple microphones 10, thereby performing processing to identify the position of a person based on the presence of vocalizations within the tunnel.

[0059] Next, in step S906, if the acoustic data processing unit 20 is able to identify the position of a person from the presence of vocalizations in the tunnel by performing voice recognition processing, it performs a correction process to add further information to identify the "person detection position" to the display data already created in step S903, and then proceeds to processing in step S907.

[0060] If the process proceeds to step S907, the acoustic data processing unit 20 executes a display process for the created display data, and ends the series of processes.

[0061] In the first embodiment described above, the monitoring area is described as the inside of a tunnel. However, an evacuation passage is provided in the tunnel. Therefore, by setting the tunnel evacuation passage as the monitoring area and installing multiple microphones 10 in the evacuation passage, it is possible to detect people in the evacuation passage.

[0062] In the first embodiment described above, a case has been described in which multiple microphones 10 are installed at both ends of the tunnel. However, the installation locations of the multiple microphones are not limited to this layout, and multiple microphones can also be arranged in the height direction, making it possible to display data based on frequency analysis results at different heights.

[0063] As described above, according to the first embodiment, display data for an operator to determine whether or not human vocalizations are present can be created and displayed based on the results of frequency analysis of acoustic data collected by multiple microphones inside a tunnel.

[0064] In particular, the tunnel monitoring system according to the first embodiment provides the following effects. Effect 1: There is a degree of freedom in the placement of the microphone, so it can be easily adapted to retrofitting or temporarily adding a function to detect people inside a tunnel. In other words, with a relatively simple configuration, it is possible to detect people who are late in escaping in the monitored area, inside a tunnel or in an evacuation route.

[0065] Effect 2: An operator who visually checks the displayed data can distinguish the presence or absence of human vocalizations from the sound data collected inside the tunnel, distinguishing it from the sound of water discharge, warning sounds, and idling vehicle sounds. In other words, it becomes easier to grasp the situation inside a tunnel, where the influence of sound reflection is significant, and it is easy to determine from the displayed data whether or not there are characteristics specific to vocalizations in the sound data collected at multiple locations inside the tunnel.

[0066] Effect 3: By further performing voice recognition processing on the acoustic data, it is possible to accurately determine whether or not a person is present in the monitored area.

[0067] Effect 4: The results of frequency analysis of the acoustic data collected by multiple microphones can identify the location of a person and also the range of the vocalization. Therefore, the operator can estimate the direction of movement of a person by visually checking the display data that is updated sequentially in chronological order.

[0068] Effect 5: By combining the human detection method according to the present disclosure with a fire detection method using other sensors, it is possible to build a system that can quickly detect people who are left behind in a tunnel in the event of a fire. [Explanation of symbols]

[0069] 10, 10(1)~10(12) Microphone (sound pickup device), 20 Acoustic data processing unit.

Claims

1. a plurality of sound collection devices arranged in a row along both ends of a tunnel in a monitoring area, the sound collection devices collecting sound data generated within the tunnel; a sound data processing unit that calculates a sound pressure level for each band by performing frequency analysis on the sound data collected by each of the plurality of sound collection devices; Equipped with The acoustic data processing unit identifies the type and location of a sound generated in the tunnel based on the result of the frequency analysis and the arrangement of the plurality of sound collection devices. Tunnel monitoring system.

2. The types of sounds include human voices in the tunnel, warning sounds, idling sounds of cars, and water spraying sounds. The acoustic data processing unit identifies, based on the result of the frequency analysis and the arrangement of the plurality of sound collection devices, the installation location of the sound collection device that obtains the highest sound pressure level in the 100 to 2000 Hz band corresponding to the frequency band of the human vocalization sound, as the human detection position. A tunnel monitoring system according to claim 1.

3. The acoustic data processing unit Calculating an average sound pressure level in a frequency band of the vocal sound for each of the plurality of sound collection devices; Identifying a sound collection device that collected sound data having a maximum value among the average sound pressure levels as a first sound collection device; Identifying a sound collection device that collects sound data having an average sound pressure level within a predetermined attenuation sound pressure amount from the maximum value as a second sound collection device, It is estimated that the voice is generated in the direction from the location where the first sound collecting device is installed to the location where the second sound collecting device is installed. A tunnel monitoring system according to claim 2.

4. The acoustic data processing unit Creating display data that allows identification of the sound pressure level for each of the bands at the respective positions of the plurality of sound collection devices, and displaying the display data on a display unit; When the first sound collecting device and the second sound collecting device can be identified, the display data is generated so as to distinguishably display the ranges of the first sound collecting device and the second sound collecting device, and the display data is displayed on the display unit. A tunnel monitoring system according to claim 3.

5. The acoustic data processing unit generates and sequentially updates the display data, and causes the display unit to display the sequentially updated display data. A tunnel monitoring system according to claim 4.

Images