JP4359496B2 - Glass breakage detector - Google Patents

Description

  The present invention relates to a glass breakage detection device that issues an alarm by detecting that a building glass or a showcase glass is broken, and uses a vibration wave and an acoustic wave generated when the glass is broken. The present invention relates to a glass breakage detection device for detecting breakage of glass.

  In recent years, damage caused by empty nests or the like that breaks down glass used for building windows and enters the interior has been increasing. There have also been many reports of damage by thieves who destroy the glass of showcases installed inside buildings and steal goods and valuables in the showcases.

  In order to prevent such damage, a glass breakage detection device that detects that glass has been broken is used. The glass breakage detection device has a function of issuing an alarm or notifying an external security center or the like via a communication line when detecting that the glass is broken.

Two types of devices that detect glass breakage are known: a type using a contact type sensor and a type using a non-contact type sensor.
As shown in Patent Document 1, the type using a contact type sensor has a strength characteristic of a high frequency component (from 100 kHz) generated when a piezoelectric vibration element is attached to a glass surface and an impact is applied to the glass. Based on this, it is determined whether or not the glass has been broken, and glass breakage is detected.

  On the other hand, the type using a non-contact type sensor detects breakage of glass by installing a microphone at a location away from glass such as a ceiling or a wall and analyzing an acoustic signal input to the microphone. For example, in Patent Document 2, based on the simultaneous occurrence of a low frequency (50 to 100 Hz) due to the deflection of the glass surface at the moment when the glass is broken and a high frequency (5 to 20 kHz) due to the glass being broken. , Detecting glass breakage.

  In Patent Document 3, sound and vibration are detected to detect the breakage (breakage, breakage) of the window glass of the automobile, and the movement of the non-detection body in the automobile after the detection is detected using a microwave or the like. The invention of an automobile safety sensor system that detects an intrusion into the automobile by detection is described.

Japanese Patent Publication No.57-57970 JP 7-700438 JP-A-6-258194

  The contact sensor described above detects glass breakage by detecting high frequency components that are the characteristics of vibration waves at the time of glass breakage, but the same high frequency component as at the time of glass breakage due to an impact that does not lead to glass breakage. There is a problem that a false alarm is issued at this time. Further, since one sensor is attached to one piece of glass and the impact applied to the glass is detected, sensors are required for the number of sheets of glass, which increases the cost. Moreover, since it sticks on a glass surface directly, it has the fault that a sensor tends to drop | omit from a glass surface. In addition, there is a problem in that the appearance is lost by sticking to the glass surface, and the position of the sensor can be easily visually recognized by a third party.

  On the other hand, the above-mentioned non-contact type sensor has the advantage that it solves the problems of the above-mentioned contact type sensor and can monitor a plurality of glasses with one sensor. There are many false alarms that cause an alarm even if the glass is not broken by reaction, and if the sensitivity of the sensor is lowered, if the sound when the glass is broken is low, the alarm will not be issued and the alarm will be lost There was a problem.

  Moreover, in the above-mentioned automobile safety sensor system, sound and vibration are detected to detect glass breakage. However, since the target is assumed to be in a quiet automobile, this is applied to a general building with a lot of environmental noise. When applied, there is a problem that a false alarm is generated only by vibration or sound due to mischief or the like.

  As described above, the conventional glass breakage detection apparatus cannot reliably determine whether or not the glass is actually broken, and cannot eliminate the problem of generating a false alarm. This is because the waveforms of vibration waves and acoustic waves generated by impacts applied to glass and environmental noise are diverse. Therefore, it is difficult to eliminate false alarms other than glass breakage with a sensor based only on vibration waves and a sensor alone based on acoustic waves, and it is not easy to deal with simply combining vibration waves and acoustic waves. .

The present invention has been made to solve the above problems, and a first object of the present invention is to provide a glass breakage detection apparatus that can reliably distinguish whether generated acoustic waves and vibration waves are due to glass breakage. It is said.
In addition, it can be securely attached to the window glass, and multiple windows can be monitored by installing a sensor in one place, and whether the generated acoustic waves and vibration waves are due to glass breakage is reliably distinguished. It is a second object to provide a glass breakage detection device that can be used.

The glass breakage detection device according to claim 1 is a glass breakage detection device that detects breakage of glass by vibration waves and acoustic waves generated when the glass is broken, and outputs vibration signals by detecting vibration waves. Glass wave breakage by determination including at least determination based on a correlation between intensity of the vibration signal and intensity of the acoustic signal, vibration wave detection means for detecting the acoustic wave and outputting an acoustic signal by detecting the acoustic wave and a determining means, before Symbol correlation, a low frequency component intensity intensity ratio of the intensity and the vibration signal of the high frequency components of the vibration signal, the intensity of a predetermined frequency component of the acoustic signal It is characterized by a correlation with a continuity representing a continuation state that is equal to or greater than a predetermined value.

The glass breakage detection device according to claim 2 is a glass breakage detection device that detects breakage of glass by vibration waves and acoustic waves generated when the glass is broken, and outputs vibration signals by detecting vibration waves. Glass wave breakage by determination including at least determination based on a correlation between intensity of the vibration signal and intensity of the acoustic signal, vibration wave detection means for detecting the acoustic wave and outputting an acoustic signal by detecting the acoustic wave Determining means for determining, wherein the correlation is such that the intensity ratio between the intensity of the high frequency component of the vibration signal and the intensity of the low frequency component of the acoustic signal and the intensity of the predetermined frequency component of the acoustic signal are predetermined. It is characterized by a correlation with the degree of continuity that represents a continuation state that is greater than or equal to the value.

Glass breaking detection apparatus according to claim 3, in the glass breaking detection apparatus according to claim 1 or 2, the high frequency band of the vibration signal is characterized in that it comprises a predetermined frequency of the acoustic signal.

The glass breakage detection device according to claim 4 is the glass breakage detection device according to any one of claims 1 to 3 , wherein the determination means indicates that the measured continuity indicates the relationship between the intensity ratio and the continuity. The occurrence of glass breakage is determined when the continuity value obtained from the discriminant function and the measured intensity ratio is exceeded.

The glass breakage detection apparatus according to claim 5 is the glass breakage detection apparatus according to any one of claims 1 to 3 , wherein the determination unit is configured such that the continuity is greater than a predetermined value and a predetermined frequency component of the acoustic signal. The continuity obtained from the discriminant function indicating the relationship between the intensity ratio and the continuity and the measured intensity ratio when the high frequency component of the acoustic signal at the time when the maximum value exceeds a predetermined value The value of the degree is compared with the measured continuity, and when the measured continuity is larger, the occurrence of glass breakage is determined.

The glass breakage detection apparatus according to claim 6 is the glass breakage detection apparatus according to any one of claims 1 to 5 , wherein the continuity of the acoustic signal is such that the intensity of a predetermined frequency component of the acoustic signal is equal to or greater than the predetermined value. It is characterized by a certain duration.

Glass breaking detection apparatus according to claim 7, in the glass breaking detection apparatus according to claims 1 to 5, continuing degree of the acoustic signal intensity of a predetermined frequency component of the acoustic signal exceeds said predetermined value It is characterized by the cumulative value of the intensity from the past.

A glass breakage detection device according to claim 8 is the glass breakage detection device according to claims 1 to 7 , wherein at least the vibration wave detection means is attached to a structural member that supports glass, and It is characterized by detecting vibration waves.

According to the glass breakage detection device described in claim 1 , not only the vibration signal and the acoustic signal are simply combined, but also at least in the correlation between the vibration signal intensity and the sound signal intensity peculiar to the event of glass breakage. Since the determination is performed based on the determination including the determination based on the above , there is an effect that it is possible to perform a reliable determination of glass breakage, particularly the strength of the high frequency component of the vibration signal and the strength of the low frequency component of the vibration signal. There is an effect that the glass breakage can be accurately determined by making a determination based on the correlation between the intensity ratio and the continuity representing the continuation state in which the intensity of the predetermined frequency component of the acoustic signal is equal to or greater than a predetermined value. is there.

According to the glass breakage detection device described in claim 2 , not only the vibration signal and the acoustic signal are simply combined, but at least the correlation between the intensity of the vibration signal and the intensity of the sound signal peculiar to the event of the glass breakage. Since the determination is based on the determination including the determination based on, there is an effect that it is possible to perform a reliable determination of glass breakage, and in particular , the strength of the high frequency component of the vibration signal and the strength of the low frequency component of the acoustic signal There is an effect that the glass breakage can be accurately determined by making a determination based on the correlation between the intensity ratio and the continuity representing the continuation state in which the intensity of the predetermined frequency component of the acoustic signal is equal to or greater than a predetermined value. is there.

According to the glass breakage detection device described in claim 3 , in the effect of the invention described in claim 1 or 2 , since an acoustic wave is generated by the vibration of the glass, there is a correlation between the two. When the high frequency band includes the predetermined frequency of the acoustic signal, a clearer correlation occurs, and there is an effect that the glass breakage can be accurately determined by performing the determination based on the correlation.

According to the glass breakage detecting device described in claim 4 , in the effect of the invention described in claims 1 to 3 , the determination means determines that the measured continuity is a relationship between the intensity ratio and the continuity. Since the occurrence of glass breakage is determined when the continuity value obtained from the discriminant function and the measured intensity ratio is exceeded, the glass breakage and others are accurately and effectively determined by the optimum discriminant function obtained by experiment. It is possible to determine the occurrence of glass breakage by identifying the above events.

According to the glass breakage detection apparatus described in claim 5 , in the effect of the invention described in claims 1 to 3 , the determination unit is configured such that the continuity is larger than a predetermined value and the predetermined frequency of the acoustic signal. When the high frequency component of the acoustic signal when the component becomes maximum exceeds a predetermined value, the discrimination function indicating the relationship between the intensity ratio and the continuity and the measured intensity ratio Compares the value of continuity with the measured continuity, and determines the occurrence of glass breakage when the measured continuity is greater, so it is efficient only for events that are likely to be glass breakage. The glass breakage can be accurately determined by the optimum discriminant function obtained through experiments.

According to the glass breakage detection device described in claim 6 , in the effect of the invention according to claims 1 to 5 , the intensity of the predetermined frequency component of the acoustic signal is not less than the predetermined value as the continuity of the acoustic signal. Since a certain duration is employed, data relating to the continuity of the acoustic signal can be obtained by simple data processing, and there is an effect that the glass breakage can be accurately determined although the burden on the apparatus is light.

According to the glass breakage detection device described in claim 7 , in the effect of the invention according to claims 1 to 5 , the intensity of the predetermined frequency component of the acoustic signal exceeds the predetermined value as the continuity of the acoustic signal. Since the cumulative value of the intensity from the past is adopted, it is possible to obtain optimum data that accurately represents the continuity of the acoustic signal, and to determine the glass breakage more accurately.

According to the glass fracture detection apparatus as claimed in claim 8, it is possible to detect the vibration wave from the structural member by attaching said vibration wave detecting means to the structural member that supports the glass, the invention described in claims 1 to 7 In the effect, there is an effect that the determination unit can determine the glass breakage by identifying the blow applied to the structural member and the glass breakage based on the correlation.

In the following, embodiments of the present invention that the patent applicant thinks best at the time of filing to implement the present invention will be described.
(1) Configuration of Embodiment FIG. 1 is a block diagram showing a configuration of a glass breakage detection apparatus 100 according to the present invention.
Reference numeral 1 denotes a vibration pickup as vibration wave detection means for detecting vibration waves, and 3 denotes a microphone as acoustic wave detection means for detecting acoustic waves. Each output is amplified to an appropriate level by the amplifiers 10 and 30. The output of the amplifier 10 is input to band pass filters (hereinafter referred to as BPF) 11 and 21. The BPF 11 is a filter having a pass characteristic of 50 Hz to 150 Hz which is a low frequency band of the vibration signal, and the BPF 21 is a filter having a pass characteristic of 3 kHz to 15 kHz which is a high frequency band of the vibration signal. Similarly, the output of the amplifier 30 is input to the BPFs 31 and 41. BPF 31 is a filter having a pass characteristic of 3 kHz to 4 kHz, which is a middle frequency band of the acoustic signal, and BPF 41 is a filter having a pass characteristic of 15 kHz to 20 kHz, which is a high frequency band of the acoustic signal.

  The outputs of the BPFs 11, 21, 31, and 41 are input to the envelope extraction circuits 12, 22, 32, and 42, respectively, and the envelope of each band signal is extracted. Although a detailed view of the envelope extraction circuit is not shown, it can be realized by a half-wave rectifier circuit or a full-wave rectifier circuit and a smoothing circuit. Outputs of the envelope extraction circuits 12, 22, 32, and 42 are respectively input to logarithmic circuits 13, 23, 33, and 43, logarithmically converted, then AD converted, and input to the CPU 5 at the subsequent stage. In the CPU 5 which is a glass breakage determination means, according to a predetermined program, it is determined whether the input signal is caused by glass breakage or caused by a cause other than glass breakage. A signal is output to the transmission unit 6, and the transmission unit 6 transmits the signal to an alarm display device (not shown) wirelessly or by wire.

  FIG. 9 is a diagram showing an attachment location when the glass breakage detection apparatus 100 is attached to the sash frame 50 that supports the glass 51 to be monitored. As shown in these drawings, it is arranged on the surface of the sash 50 that supports the glass 51 to be monitored or inside the sash 50. In addition, when the sash 50 and the sash 60 are coupled by locking the crescent lock 52, the glass 61 supported by the sash 60 can be monitored.

  FIG. 10 is a perspective view showing an example of an attachment state of the vibration pickup 1 and the microphone 3 of the glass breakage detection apparatus 100. The vibration pickup 1 needs to be arranged so as to be in direct contact with the sash frame 50, but the microphone 3 is optional as long as it is in the vicinity of the glasses 51 and 61 so that the glasses 51 and 61 can be monitored simultaneously. For example, in FIG. 10, the vibration pickup 1 and the microphone 3 are housed and integrated in a common housing as the glass breakage detection device 100, but only the vibration pickup 1 is disposed so as to be in direct contact with the sash frame 50. The microphone 3 can monitor the glass 51 and 61 at the same time, but can be arranged as a separate device at a position away from the sash 50 and 60.

(2) Outline of Glass Breakage Detection in Embodiment The principle of glass breakage detection according to the present invention realized in the embodiment will be specifically described with reference to FIGS.
The main object of the present invention is to prevent false alarms that occur as much as possible when a hard object is hit with a sash frame, or when pebbles raised by an automobile or the like collide with glass to such an extent that they do not break. It is to eliminate. The inventors of the present invention found that the vibration wave generated in the sash frame when the glass is broken is hard while the low-frequency (50 Hz to 150 Hz) and high-frequency (3 kHz to 15 kHz) components are equivalent in strength. If the sash frame is struck or a pebbles collides with the glass, the high frequency component of the vibration wave is generated but the low frequency component of the vibration wave is small, and if the glass is broken, the acoustic wave Experiments have found that the continuity of components in the middle range of the acoustic wave is high, such as the continuity of the power envelope signal in the middle range (3 to 4 kHz band).

  For example, as shown in FIG. 5, the continuity of the power envelope signal is an accumulation of SM (n) −Thd2 in a section where the power envelope signal SM (n) exceeds a predetermined threshold Thd1 and falls below a threshold Thd2 smaller than Thd1. It can be represented by the value S. The shaded area in FIG. 5 represents the process of calculating the cumulative value up to the current time n.

  From the above, when the glass is broken, the intensity of the high-frequency component of the vibration wave and the intensity of the low-frequency component of the vibration wave are in the same or no significant difference, so the high-frequency component of the vibration wave and the low-frequency component of the vibration wave The intensity ratio (intensity ratio) of each component is close to 1 or a value before and after that, in other words, the difference Vdiff between the high frequency component of the vibration wave and the low frequency component of the vibration wave in the logarithmic value is smaller than a predetermined value. . In this embodiment, since the envelope signals of the acoustic wave and the vibration wave are logarithmized, the difference between the high frequency component and the low frequency component is set to Vdiff. It is clear that

  In addition, when the glass is broken, it is considered that the intensity of the predetermined frequency component of the acoustic signal is maintained at a predetermined value or more within a certain time, and thus the continuation indicating the degree of the continuous state where the intensity is at or above the predetermined value. Considering the accumulated value S as a degree, it can be considered that the glass is broken when the accumulated value S is larger than another predetermined value.

  However, in actuality, it is not possible to judge glass breakage with a simple combination of these two values, and as explained below, there is a positive correlation that the cumulative value S increases as the difference Vdiff increases. Was discovered by the inventors of the present application, and the present invention experimentally obtained a discriminant function S = a × Vdiff + b (a> 0) using the correlation between the two values, and used this to determine the presence or absence of glass breakage. This is one of the important points, and this has achieved a great effect.

A specific example of the correlation is shown in FIG.
FIG. 4 schematically shows data obtained from experiments conducted by the inventors of the present application. The data at the time of breaking the glass, and when the sash frame was struck or when pebbles collided with the glass, it was destroyed. The data when not reached (events other than glass breakage) are shown in the figure with the horizontal axis representing the difference Vdiff and the vertical axis representing the cumulative value S.

  Here, if there is no correlation between the difference Vdiff and the accumulated value S, there should be no fixed relationship in relation to each variable on the horizontal axis of the vertical axis at the plot position of the data in FIG. For example, when a sash frame is struck and pebbles collide with the glass but it does not break (events other than glass breakage), the data should be sparsely present in the upper right, lower right, and lower left areas in the figure. is there.

  However, since there is a positive correlation between the difference Vdiff and the cumulative value S, the data of events other than glass breakage (represented by ◯ in FIG. 4) is actually shown by the oblique broken lines in FIG. The data distribution is ascending to the right, and the data due to glass breakage (represented by □ in FIG. 4) is also increasing to the right.

  Here, for the purpose of discriminating between glass breakage and other events, as means for dividing the data at the time of glass breakage and the data at the time of other event occurrence on the graph of FIG. An identification line as indicated by a solid line and an identification function S = a × Vdiff + b (a> 0) representing this identification line are introduced. Here, the gradient a of the discrimination function is set so as to be substantially parallel to the gradient of the data distribution in FIG. That is, the gradient a of the discrimination function represents the gradient corresponding to the correlation, and this value can vary depending on the characteristics of the glass and the installation environment. ing. The intercept b of the discriminant function is a value indicating to which data side the discriminant line that separates the glass breakage data and other event data is set, and it is desirable to reliably detect the glass breakage. If you want to suppress the false alarm more, you can set it to a large value.

Hereinafter, the correlation between the difference Vdiff and the accumulated value S will be described in more detail.
The acoustic wave generated by hitting or breaking the glass or sash frame, which is the subject of the present application, is induced as a result of the formation of density in the air by the vibration wave of the vibrated glass or sash. Therefore, a high-frequency vibration wave can generate a high-frequency rough wave (acoustic wave), and a low-frequency vibration wave can generate a low-frequency rough wave (acoustic wave). A strong vibration wave can generate a strong acoustic wave, and a weak vibration wave can generate a weak acoustic wave. Therefore, an event that generates a high frequency component (3 kHz to 15 kHz) of a strong vibration wave induces a strong acoustic wave of 3 to 4 kHz included in this frequency band, and conversely, a high frequency component of a weak vibration wave. The event that generates (3 kHz to 15 kHz) induces only weak acoustic waves in the 3 to 4 kHz band.

  On the other hand, since the accumulated value S is the accumulation of the difference between the envelope of the 3 to 4 kHz component of the acoustic wave and the threshold value Thd2, the stronger the 3 to 4 kHz component of the acoustic wave, the greater the S.

  In summary, an event that generates a high frequency component (3 kHz to 15 kHz) of a strong vibration wave induces a strong acoustic wave in the 3 to 4 kHz band, and as a result, the accumulated value S increases. In addition, when the strength of the high frequency component of the vibration wave increases, it means the ratio between this value and the strength of the low frequency component of the vibration wave, and Vdiff defined by the difference when the intensity is logarithmically increased. Therefore, a positive correlation is obtained between Vdiff and S as a final result.

  As described above, the glass breakage detection apparatus according to the present invention has the intensity ratio (difference between logarithmic values) of the high frequency component of the vibration wave and the low frequency component of the vibration wave, and the continuation of the power envelope signal in the middle range of the acoustic wave. Further, by utilizing the correlation between these two values, events other than glass breakage and glass breakage are distinguished and glass breakage is effectively detected.

  In addition, vibration waves have longitudinal wave components, transverse wave components, surface wave components, etc., and not all of these contribute to the generation of acoustic waves, so acoustic waves and vibration waves do not provide exactly the same information. Actually, the continuity of the signal is not found in the vibration wave, and the point that the cumulative value S is calculated from the acoustic wave is also an important feature of the present invention.

(3) Specific Procedure for Information Processing for Glass Breakage Detection in Embodiment Next, a specific procedure for information processing for glass breakage detection in the present invention is executed inside the CPU 5 shown in FIG. The processing flow will be described with reference to FIG.
Each output of the logarithmic circuits 13, 23, 33 and 43 in FIG. 1 is a logarithmized envelope signal, which is sampled and input to the CPU 5. In FIG. 2, each sampled input signal is represented by a low-frequency vibration signal VL (n), a high-frequency vibration signal VH (n), a mid-frequency acoustic signal SM (n), and a high-frequency acoustic signal SH (n ). n means a temporal sampling point of each signal, and the sampling period is preferably a sampling period that does not drop the logarithmic envelope characteristic information amount such as 10 ms.

  In step 302, it is determined whether the mid-frequency acoustic signal SM (n), which is a logarithmized envelope signal in the mid-range (3 to 4 kHz) of the acoustic wave, is equal to or greater than a predetermined value (Thd1). If it is equal to or greater than Thd1, it is determined that a signal due to an event that may cause glass breakage has been input, and processing for determining whether or not glass breakage is started. If it is equal to or less than Thd1, the process returns to step 302 to continue monitoring the input signal.

  Thd1 is a value determined by the characteristics of each circuit such as the sensitivity of the microphone, the amplification factor of the amplifier, and the average background noise level of the environment in which the present glass breakage detection apparatus is used. The logarithm shown in FIG. When the conversion circuit is a circuit that performs 20 log (X) conversion on the input X, it becomes a unit of [dB], for example, 70 [dB].

  In step 304, the event start detection flag is initialized to FALSE and S is initialized to 0. As will be described in detail later, the event start detection flag is a flag indicating that the start of an event to be detected / judged including glass breakage is detected, and S is a continuation state in which the intensity of the acoustic wave is a predetermined value or more. Is an index indicating the cumulative value of the mid-frequency acoustic signal SM (n), which is an envelope signal in the mid-range of the acoustic wave.

  In step 306, it is determined whether the mid-frequency acoustic signal SM (n) is less than a predetermined value (Thd2). Thd2 is a value determined by the characteristics of each circuit such as the sensitivity of the microphone, the amplification factor of the amplifier, and the background noise level of the environment in which the present glass breakage detection apparatus is used, for example, 60, and Thd2 is Thd1. The value needs to be about 10 dB smaller.

  The relationship between Thd1 and Thd2 is shown in FIG. If the mid-frequency acoustic signal SM (n) is less than Thd2, the process returns to step 302 assuming that the signal due to the event including glass breakage has ended. If it is equal to or greater than Thd2, the process proceeds to step 308 assuming that the signal continues.

  In step 308, the value (SM (n) -Thd2) of the mid-frequency acoustic signal SM (n) exceeding Thd2 is cumulatively added to S. In FIG. 5, n means the current sampling point, and the shaded area means the value S accumulated up to the sampling point.

  In step 310, it is determined whether or not the event start detection flag is TRUE. Since it is FALSE in the initial state, the process proceeds to step 312, but when the predetermined conditions of steps 312 and 314 are already satisfied, the event start detection Flag is set to TRUE and the process proceeds to step 320.

  In step 312, it is determined whether the mid-frequency acoustic signal SM (n), which is an envelope signal, has a maximum value in the range of SM (n-10) to SM (n + 10). In this step, the maximum value of the mid-frequency acoustic signal SM (n) is determined at the moment when an event such as an object hitting a structure (frame) supporting the glass or the like occurs or the glass breaks. The frequency acoustic signal SM (n) is considered to be maximized, and the intensity ratio of the low-frequency vibration signal VL (n) and the high-frequency vibration signal VH (n) at this time is calculated in a later step 318. This is to make a judgment material for identifying an event. In order to determine the maximum value between SM (n-10) and SM (n + 10), it is necessary to prepare a sufficient memory in the CPU 5 to support the value of SM (n). If SM (n) is the maximum value, the process proceeds to step 314. If SM (n) is not the maximum value, the process returns to step 306.

  In step 314, it is determined whether the high frequency vibration signal VH (n), which is an envelope signal, exceeds a predetermined value Thd3. This is based on the fact that high-frequency vibration waves are observed in the sash frame when the glass is broken, and noise that does not cause vibration of the sash that occurs in daily life, such as accidentally breaking tableware. This is to eliminate the output of a false alarm. Thd3 is preferably determined based on the vibration intensity at the time of glass breakage, and is set to 90, for example. If the high frequency vibration signal VH (n) is larger than Thd3, the glass breakage determination is continued and the process proceeds to step 318. If it is equal to or less than Thd3, the process returns to step 306.

  The conditions in steps 312 and 314 are for detecting the start of an event to be detected and judged including glass breakage. When these conditions are satisfied, in step 318, the event start detection Flag is set to TRUE. . In addition, the characteristics observed at the early stage of the event are effective in identifying and judging whether the event is a glass breakage or other events, and support that these values can be used in the subsequent steps.

  That is, the intensity ratio between the low frequency vibration signal VL (n) and the high frequency vibration signal VH (n) at the time point n when the mid frequency acoustic signal SM (n) becomes maximum is calculated in the form of a logarithmic difference. Then, Vdiff = VH (n) −VL (n) is set, and the high-frequency acoustic signal SH (n) at the simultaneous point n is set as SH = SH (n) as the high-frequency acoustic signal SH.

  In step 320, it is determined whether or not the cumulative value S calculated in step 308 is larger than a predetermined value Thd5 and SH is larger than a predetermined value Thd6. If the accumulated value S and the high frequency acoustic signal SH are both larger than the reference value, the process proceeds to step 322 because it has a glass breaking characteristic. If either one is smaller than the reference value, the process returns to step 306, S is updated in step 308, and this determination loop is repeated. The predetermined value Thd5 is, for example, 220, Thd6 is 70, etc., and these values are determined by a glass breaking experiment.

  Step 320 is based on the feature that when the glass is broken, both S and SH tend to increase, but when the glass is hit but not destroyed, these numbers are small. This step can eliminate glass blows that do not cause breakage. That is, according to the judgment in step 320, the glass breakage and the glass blow that does not lead to breakage are identified from the events including the glass breakage detected in step 318. It is possible to prevent an erroneous report from being output when there is no such information. The relationship between S and SH at step 320 is shown in FIG. FIG. 3 schematically shows data obtained from experiments by the inventors.

  In step 322, it is determined whether or not the cumulative value S calculated in step 308 is larger than a value F (Vdiff) = a × Vdiff + b, (a> 0) determined by Vdiff. a × Vdiff + b is called a discriminant function and is a straight line in this example, but there is no particular limitation on the order of the formula, and glass breakage and other events can be distinguished based on the correlation between Vdiff and S Any function can be used. F (Vdiff) = a × Vdiff + b, (a> 0) is, for example, 10Vdiff + 80 or the like, which is determined by a glass breaking experiment.

  If the accumulated value S is larger than the above-described F (Vdiff), it is determined that the accumulated value S has a glass breaking characteristic, and the process proceeds to step 324. If it is smaller, the process returns to step 306, S is updated in step 308, and then the determination loop is repeated.

  This step is hard and is intended to eliminate false alarms when a sash frame is hit or when pebbles hit the glass. As shown in FIG. 4, the vibration wave when the glass is broken includes components from a low frequency region to a high frequency region. Therefore, the difference between the high frequency component and the low frequency component of the vibration wave, that is, Vdiff = VH−VL. Is a value around 0, but when the sash frame is hard and is hit or when it collides with glass such as pebbles, high frequency components of vibration waves are generated, but low frequency components are small Therefore, Vdiff is a large value.

  Further, as described in step 320, since S tends to increase when the glass breaks, it can be considered that the glass is broken when Vdiff is smaller than a certain predetermined value and S is larger than another predetermined value.

  However, in actuality, rather than a simple combination, the discriminant function F (Vdiff) = a × Vdiff + b, (a> 0) using a positive correlation that S increases as Vdiff increases as described above is used. With very effective results. Further, if necessary, conditions such as the absolute value of Vdiff being smaller than a predetermined value or S being larger than another predetermined value can be used in combination, and the glass breakage detection accuracy can be further increased. .

In step 324, if all the above conditions are satisfied, it is determined that the glass is finally broken, and an alarm is output.
Further, in the course of the loop that repeats steps 308, 310, 320, and 322 from step 320 to step 306, or from step 322 to step 306 and returning to step 308, the mid-frequency is reached before satisfying all the conditions for glass breakage. When the acoustic signal SM (n) becomes smaller than Thd2, the event that is the cause of the input signal is not glass breakage, and the process returns to step 302, which is the head of the present process.
In the above description, the loop returning to step 306 and the loop returning to step 302 are performed in synchronism with the period in which the CPU 5 fetches the next sampling point.

  In step 308, the value (SM (n) -Thd2) of the mid-frequency acoustic signal SM (n) that exceeds Thd2 is cumulatively added to S. In steps 320 and 322, whether or not glass breakage occurs based on the cumulative value S. However, as the continuity indicating the degree of the continuation state in which the intensity of the predetermined frequency component of the acoustic signal maintains a predetermined value or higher, the current sampling point n is exceeded after exceeding Thd1 instead of the cumulative value S. (Time duration T) can also be used (see FIG. 5).

  FIG. 6 shows a flowchart when the duration T is used as the degree of continuity instead of the cumulative value S. The changed steps are 304, 308, 320, 322, and the corresponding steps are 304 ', 308', 320 ', 322', respectively. The change points are steps 304 ′, 320 ′, and 322 ′, in which S is changed to T, and in steps 320 ′ and 322 ′, in addition, the threshold Thd5 is set to Thd5 ′, and the parameters a and b for determination are Since only the points a ′ and b ′ are changed and do not deviate from the gist described in FIG. 2, the detailed description is omitted. Step 308 'is a process for calculating the time from when SM (n) exceeds Thd1 to the current sampling point as shown in FIG.

  In the embodiment shown in FIGS. 1 and 2, the difference between these logarithmic values Vdiff = VH (n) −VL (n) is calculated as the intensity ratio between the high frequency vibration signal and the low frequency vibration signal. However, the same effect can be obtained by using a low frequency acoustic signal (low frequency component of acoustic wave) instead of the low frequency vibration signal.

  FIG. 7 shows a configuration block diagram in that case. The difference from FIG. 1 is that blocks 51, 52 for acquiring a low frequency acoustic signal (low frequency component of an acoustic wave) instead of the blocks 11, 12, 13 for acquiring the low frequency component of the vibration wave. 53. The output of the amplifier 30 is also input to the BPF 51. The BPF 51 is a filter having a pass characteristic of 50 Hz to 100 Hz. The output of the BPF 51 is input to the envelope extraction circuit 52, and the envelope is extracted. The output of the envelope extraction circuit 52 is input to the logarithmic circuit 53, logarithmically converted and output, input to the CPU 5 at the subsequent stage, A / D converted by the A / D conversion means, and processed. Since components other than the configurations related to reference numerals 51, 52, and 53 are the same as those in FIG. 1, detailed description thereof is omitted.

  FIG. 8 shows a flowchart of processing executed in the CPU 5. The difference from FIG. 2 is that the output signals of the logarithmic circuits 23, 33, 43, 53 sampled and input to the CPU 5 are the high frequency vibration signal VH (n), the mid frequency acoustic signal SM (n), For the discrimination determined from the point expressed as the low frequency acoustic signal SH (n), the low frequency acoustic signal SL (n), the Vdiff setting method in step 318 ″, and the correlation between Vdiff and S in step 322 ″. Since these are only parameters a ″ and b ″, a detailed description thereof will be omitted.

  In the glass breakage detection apparatus, the BPF is used. However, the low pass filter can be used within a range in which low frequency components, high frequency component vibration waves and acoustic waves generated due to glass breakage can be measured. Alternatively, a configuration using a high-pass filter may be used. In addition, it is good also as a structure which implement | achieves the function equivalent to BPF comprised by the hardware in the said embodiment, an envelope extraction circuit, and a logarithmization circuit with the software of CPU5. On the other hand, an analog circuit including a comparator or the like instead of the CPU 5 may be used to realize a function equivalent to that shown in FIG.

In the embodiment described above, since the vibration pickup 1 and the microphone 3 are provided in the sash frame, the following effects can be obtained particularly in connection with the prior art.
That is, an existing contact type glass sensor in which a sensor is directly attached to glass monitors a very high frequency of 100 kHz or more. However, a vibration wave having a frequency of 100 kHz or more is not generated to the extent that the glass or the frame is hit normally. Therefore, according to the existing contact-type glass sensor, since the alarm is not issued when the glass is hit with a pebble or when the glass or the frame is hit, the possibility that the impact on the window frame becomes a false alarm is low.

  However, if a structure in which a sensor (at least the vibration pickup 1) is installed in the window frame in order to obtain the advantage of monitoring two or more pieces of glass with one device as in the embodiment of the present application, this time, 100 kHz The high frequency vibration wave is not transmitted to the window frame. This is because there are many boundary surfaces between the glass and the sensor, such as a wound rubber, a sealing agent, and a frame for fixing the glass, and 100 kHz vibration passes and disappears.

  Therefore, when installing a sensor on a window frame, the only way to detect breakage is at a frequency lower than 100 kHz, for example, a vibration wave of about 10 or more kHz, but this time, the window frame will not be directly hit or the glass will not break. There can be assumed a problem that the sensor detects a vibration of about several tens of kHz due to an impact of a certain degree. Further, as shown in FIG. 4, simply by combining the determination by the acoustic wave and the determination by the vibration wave, it is impossible to reliably separate the glass breakage from the other, and the problem cannot be avoided.

  However, according to the above-described embodiment of the present invention, since the sensor is installed on the window frame in order to monitor two or more pieces of glass with one device, the glass breakage is caused by a high-frequency vibration wave of 100 kHz or more. Despite the fact that the glass cannot be monitored, it has been found that there is a certain correlation between the intensity of the acoustic wave and the vibration wave that occurs when the glass is broken. Since it is clearly distinguishable from other events, it is possible to accurately determine the occurrence of glass breakage in an apparatus that is economical and compact for two or more glasses. can get. The invention of the present application can achieve the above effects in the above-described embodiment. However, even if the sensor is not installed in the window frame, the above-mentioned correlation can be used as long as the glass vibration is transmitted (glass surface, etc.). It goes without saying that can be used.

  Therefore, according to the present invention, even if a vibration wave having a high frequency equivalent to glass breakage is generated due to some event other than glass breakage, not only such vibration wave but also vibration wave and acoustic wave are generated. Since the determination is made using a certain correlation between the intensity of the acoustic wave and the vibration wave, which is considered to occur only when the glass is broken, the glass breakage is regarded as other events. The effect that it can be clearly identified is obtained.

It is the block diagram which showed the structure of embodiment in this invention. It is the flowchart (flow diagram) which showed the processing flow of the glass breakage determination in the glass breakage detection apparatus of embodiment. In embodiment, it is a figure which shows the graph which distinguishes the glass damage by the value of the accumulated value S and the high frequency sound signal SH, and the glass hit | damage which does not lead to a destruction. In embodiment, it is a figure which shows the graph which distinguishes the glass fracture | rupture by the correlation of the accumulated value S and the difference Vdiff, and other events. In the embodiment, it is a graph illustrating the change of the mid-range acoustic signal on the vertical axis with respect to time n on the horizontal axis, and shows the meaning of the accumulated value S and the calculation method thereof. It is the flowchart (flow chart) which showed the modification which used the accumulated value S as continuation time as time T in embodiment. It is a block diagram which shows the modification which used the low frequency acoustic signal (low frequency component of the acoustic wave) instead of the low frequency vibration signal in the embodiment. It is the flowchart (flow diagram) which showed the processing flow of the glass break determination in the glass break detection apparatus of the modification of FIG. It is the schematic which shows the installation state of the glass breakage detection apparatus of embodiment. It is an expansion perspective view which shows the installation state of the glass breakage detection apparatus of embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Vibration pickup as vibration wave detection means 2 ... Microphone (microphone) as acoustic wave detection means
5 ... CPU as determination means
51, 61 ... Glass 100 ... Glass breakage detector VL (n) ... Low frequency vibration signal,
VH (n) ... high frequency vibration signal,
SL (n): low frequency acoustic signal,
SM (n): mid frequency acoustic signal,
SH (n) ... high frequency acoustic signal,
S: Cumulative value as continuity of acoustic signal,
T: Duration as the duration of the acoustic signal,

Claims (8)

In a glass breakage detection device for detecting breakage of glass by vibration waves and acoustic waves generated when the glass is broken,
Vibration wave detecting means for detecting a vibration wave and outputting a vibration signal;
Acoustic wave detecting means for detecting an acoustic wave and outputting an acoustic signal;
Determination means for determining glass breakage by determination including at least determination based on a correlation between the intensity of the vibration signal and the intensity of the acoustic signal;
The correlation represents an intensity ratio between the intensity of the high frequency component of the vibration signal and the intensity of the low frequency component of the vibration signal, and a continuation state in which the intensity of the predetermined frequency component of the acoustic signal is greater than or equal to a predetermined value. A glass breakage detection device characterized by a correlation with continuity. In a glass breakage detection device for detecting breakage of glass by vibration waves and acoustic waves generated when the glass is broken,
Vibration wave detecting means for detecting a vibration wave and outputting a vibration signal;
Acoustic wave detecting means for detecting an acoustic wave and outputting an acoustic signal;
Determination means for determining glass breakage by determination including at least determination based on a correlation between the intensity of the vibration signal and the intensity of the acoustic signal;
The correlation represents an intensity ratio between the intensity of the high frequency component of the vibration signal and the intensity of the low frequency component of the acoustic signal, and a continuation state in which the intensity of the predetermined frequency component of the acoustic signal is greater than or equal to a predetermined value. A glass breakage detection device characterized by a correlation with continuity. The glass breakage detection apparatus according to claim 1 or 2 , wherein the high frequency band of the vibration signal includes a predetermined frequency of the acoustic signal. The determination means determines that the glass breakage occurs when the measured continuity exceeds the discriminant function indicating the relationship between the intensity ratio and the continuity and the continuity value obtained from the measured intensity ratio. glass breaking detection apparatus according to claim 1 to 3 judges. The determination means determines the intensity when the high frequency component of the acoustic signal exceeds a predetermined value when the continuity is greater than a predetermined value and the predetermined frequency component of the acoustic signal is maximized. When the measured continuity is greater than the value of the continuity obtained from the discriminant function indicating the relationship between the ratio and the continuity and the measured intensity ratio, the glass breakage is observed. glass breaking detection apparatus according to claims 1 to 3 determines the occurrence of. The continuation of the acoustic signal, the glass breaking detection apparatus according to claims 1 to 5 intensity in a predetermined frequency component of the acoustic signal is at a duration said predetermined value or more. The continuation of the acoustic signal, the glass breaking detection apparatus according to claims 1 to 5 intensity in a predetermined frequency component is the cumulative value of the intensity from exceeding the predetermined value of the acoustic signal. The glass breakage detection apparatus according to claim 1, wherein at least the vibration wave detecting means is attached to a structural member that supports glass, and the vibration wave is detected from the structural member. apparatus.

Images