JP2013032929A - Vibration detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a vibration detection method which is less likely to generate erroneous determination, even when electromagnetic noise of which frequency or time variation is similar to a reflected wave from a battery-less vibration sensor, modulated by a vibration signal, is mixed.SOLUTION: The vibration detection method uses a battery-less vibration sensor having a single port surface acoustic wave resonator and an interrogator, and detects a vibration phenomenon, on the basis of change generated at a reflected wave from the battery-less vibration sensor, according to mechanical vibration applied to the battery-less vibration sensor. A correlation coefficient, between a second data stream obtained by performing IFFT operation on an absolute value of a first data stream obtained by FFT operation of the reflected wave, and a beforehand stored reference data stream to be a standard for the vibration to be detected, is obtained, and the correlation coefficient is compared to a predetermined threshold value of the correlation coefficient, to determine presence or absence of the vibration to be detected.

Description

  The present invention relates to a vibration detection method for detecting vibration of an object using a batteryless vibration sensor including a single-port surface acoustic wave resonator, and more particularly to a band of several tens to several hundreds of kHz accompanying an object destruction phenomenon. The present invention relates to a vibration detection method for detecting AE (Acoustic Emission) waves.

  The vibration sensor is used for a seismic diagnosis of a building, a glass breakage detection device for crime prevention security, or an abnormal vibration detection of equipment or a machine tool.

  Conventionally, Patent Document 1 discloses a vibration detection system and a vibration sensor that can detect instantaneous vibration such as vibration of an AE wave generated in a breaking phenomenon of glass or the like. The vibration detection method described in Patent Document 1 is a vibration detection method in which an AE wave generated due to destruction or deformation of an object is detected by a batteryless vibration sensor including a single-port surface acoustic wave resonator. This vibration detection method applies RFID (Radio Frequency IDentification) technology. A carrier wave is sent from the interrogator and the change of the reflected wave from the single-port surface acoustic wave resonator of the vibration sensor attached to the object is monitored. Based on the change that occurs in the reflected wave due to the vibration phenomenon of the object This is a method for detecting a vibration phenomenon of an object. This type of vibration detection method does not require a power source or a signal processing circuit in the vibration sensor portion, and thus is suitable for reducing the size and height of the vibration sensor.

Japanese Patent No. 4345984

  However, in the prior art, when electromagnetic noise similar to the frequency or time change of the vibration phenomenon to be detected is mixed, it is difficult to distinguish them. For example, in the vibration detection method of Patent Document 1, a reflected wave modulated by a vibration signal is monitored with respect to a carrier wave transmitted from an interrogator. Because it is a method for determining the presence or absence of vibration events, there is a problem that erroneous determination is likely to occur if electromagnetic noise emitted from radio equipment or electronic equipment existing in the same frequency band as the carrier wave interferes with the interrogator via the antenna There is. In particular, a problem arises when electromagnetic waves having a frequency and time change similar to those of a reflected wave from a batteryless vibration sensor modulated by a vibration signal are mixed.

  SUMMARY OF THE INVENTION Accordingly, an object of the present invention is to provide a vibration detection method that is less prone to erroneous determination even when a reflected wave from a batteryless vibration sensor modulated by a vibration signal is mixed with electromagnetic noise whose frequency and time change are similar. There is.

  In order to solve the above problems, a vibration detection method according to the present invention uses a batteryless vibration sensor and an interrogator, and the batteryless vibration sensor is formed on the piezoelectric substrate and the piezoelectric substrate. A single-port surface acoustic wave resonator having a comb-like electrode; and a support substrate that supports the piezoelectric substrate, the piezoelectric substrate being supported by the support substrate in a cantilever shape. The quality factor Q value of the cantilever vibrator is 50 to 3,000, the batteryless vibration sensor is attached to an object, and the interrogator continuously transmits to the batteryless vibration sensor. While transmitting a wave, the reflected wave from the single-port surface acoustic wave resonator of the batteryless vibration sensor is monitored, and the batteryless vibration sensor is detected along with the vibration phenomenon of the object. A vibration detection method for detecting the vibration phenomenon based on a change generated in the reflected wave according to a mechanical vibration applied to the first wave, and is calculated by an FFT (Fast Fourier Transform) operation of the reflected wave. A second data string obtained by calculating an absolute value of one data string by IFFT (Inverse Fast Fourier Transform), and a reference data string that is stored in advance and serves as a reference for vibration to be detected A correlation coefficient is obtained, and the presence or absence of vibration to be detected is determined by comparing the correlation coefficient with a predetermined correlation coefficient threshold value.

  As described above, the basic structure of the batteryless vibration sensor used in the vibration detection method of the present invention is the same as that of the conventional batteryless vibration sensor described in Patent Document 1, but in the present invention, a piece by a piezoelectric substrate is used. The quality factor Q value of the cantilever vibrator is set to 50 or more so that practical sensitivity can be obtained, and is set to 3,000 or less so that reproducibility in manufacturing can be easily obtained. Furthermore, in the present invention, as in the prior art, instead of directly detecting vibration from the waveform of the reflected wave from the single-port surface acoustic wave resonator and determining it, the reference waveform is processed after calculation processing. Judgment is made by correlating with the waveform. That is, the correlation coefficient between the second data string obtained by IFFT calculation of the absolute value of the first data string calculated by the FFT calculation of the reflected wave and the pre-stored reference data string is obtained. The presence or absence of vibration is determined by comparing the correlation coefficient with a predetermined threshold value.

  By performing the above FFT and IFFT processing, the characteristic of the waveform of the reflected wave input to the interrogator is clearly extracted. By comparing it with the reference waveform of the vibration to be detected, It is possible to discriminate whether the detected signal is a reflected wave due to vibration to be detected or electromagnetic noise.

  As described above, the present invention provides a vibration detection method that is unlikely to cause erroneous determination even when electromagnetic noise having a frequency and time change similar to a reflected wave from a batteryless vibration sensor modulated by a vibration signal is mixed. It is done.

The perspective view which shows typically the structure of the battery-less vibration sensor used for one Embodiment of the vibration detection method by this invention. The system block diagram of the vibration detection system which consists of a battery-less vibration sensor and interrogator used for one Embodiment of the vibration detection method by this invention. The block block diagram of the interrogator used for one Embodiment of the vibration detection method by this invention. The flowchart figure of the signal processing in one Embodiment of the vibration detection method by this invention. It is a figure which shows an example of the vibration waveform of the vibration used as detection object, and is a figure which shows the destruction signal by the destruction vibration produced | generated when glass is broken. The figure which shows an example of the output signal from the band pass filter of an interrogator when an interference wave exists, ie, a time response waveform. The figure which shows an example of the 1st data sequence which performed the FFT calculation process of the output signal from the band pass filter of an interrogator. The figure which shows an example of a 2nd data sequence. The figure which shows an example of the correlation between a reference | standard data string and a 2nd data string.

  Embodiments of the present invention will be described below in detail with reference to the drawings.

  FIG. 1 is a perspective view schematically showing the structure of a batteryless vibration sensor used in an embodiment of a vibration detection method according to the present invention. In FIG. 1, a batteryless vibration sensor 1 includes a single-port surface acoustic wave resonator having a piezoelectric substrate 4 and comb-shaped electrodes 3 formed on the piezoelectric substrate 4, and a support substrate that supports the piezoelectric substrate 4. The piezoelectric substrate 4 is supported by the support substrate 5 in a cantilever shape to form a cantilever vibrator, and the quality factor Q value of the cantilever vibrator is 50 to 3,000. ing. Further, the batteryless vibration sensor 1 includes an antenna 2 connected to a comb electrode 3. In the batteryless vibration sensor 1, when vibration is applied to the piezoelectric substrate 4 from the outside, the piezoelectric substrate 4 is distorted, and the distance between the electrode fingers constituting the comb electrode 3 and the elastic wave propagating through the piezoelectric substrate 4 are generated. The speed changes, and as a result, the impedance of the single-port surface acoustic wave resonator changes. Here, the resonance frequency of the cantilever vibrator configured by supporting the piezoelectric substrate 4 on the support substrate 5 is designed to belong to a characteristic frequency band in the mechanical vibration of the object to be detected. Yes.

  FIG. 2 is a system configuration diagram of a vibration detection system including a batteryless vibration sensor and an interrogator used in an embodiment of the vibration detection method according to the present invention. In FIG. 2, as the batteryless vibration sensor 11, for example, the batteryless vibration sensor 1 shown in FIG. 1 can be used. The batteryless vibration sensor 11 is attached to the object, and the interrogator 12 sends a continuous transmission wave to the batteryless vibration sensor 11, while the reflected wave from the single-port surface acoustic wave resonator of the batteryless vibration sensor 11. Is detected, and the vibration phenomenon of the object is detected based on the change generated in the reflected wave in accordance with the mechanical vibration given to the batteryless vibration sensor 11 in accordance with the vibration phenomenon of the object. That is, the antenna of the batteryless vibration sensor 11 receives the transmission wave 13 radiated from the interrogator 12 and radiates the reflected wave 14 corresponding to the impedance change of the surface acoustic wave resonator. The reflected wave 14 radiated at this time has a waveform in which the carrier wave of the transmission wave 13 is modulated by external vibration. In addition to the reflected wave 14, an interfering wave 15 such as electromagnetic noise may be input to the antenna of the interrogator 12.

  FIG. 3 is a block diagram of an interrogator used in one embodiment of the vibration detection method according to the present invention. In FIG. 3, a transmission frequency signal is generated by a transmission unit 24 composed of a known PLL (Phase Locked Loop), etc., and is demultiplexed into two by a demultiplexer 25, and one is transmitted to a power amplifier 26 as a transmission signal. The other is sent to the mixer 21 via the attenuator 23 as a local transmission signal. The transmission signal is transmitted from the antenna 28 via the circulator 27 with a desired amount of power. The transmitted wave is sent to the batteryless vibration sensor. A weak reflected wave from the batteryless vibration sensor with respect to the transmitted wave is received from the antenna 28 and amplified by the low noise amplifier 20 via the circulator 27. Thereafter, the signal is down-converted by the mixer 21, and an output signal having a vibration signal is taken out by the subsequent band-pass filter 22. The output signal is input to the signal processing unit 29, and signal processing as described below is performed.

  FIG. 4 is a flowchart of signal processing in one embodiment of the vibration detection method according to the present invention. In FIG. 4, first, the intensity of the output signal that has passed through the bandpass filter 22 of the interrogator of FIG. 3 is not less than a preset threshold value, and the duration of the signal is not less than a predetermined time. A data group of the signal part is extracted. Next, an FFT operation is performed on the data group every desired frame time to create a first data string. After that, the absolute value of the first data string is calculated, the IFFT operation is executed, and the second data string is created. Finally, the second data string is compared with a reference data string stored in advance as a standard vibration waveform of the AE wave due to the destruction to be detected, and a correlation coefficient between these data strings is determined in advance. If the correlation coefficient is equal to or greater than the threshold value, it is determined that a destruction signal has been detected, and the detection result is output.

  In the present invention, when a reflected wave including a destruction signal is input, a strong correlation is observed between the second data sequence and the reference data sequence, and when an interference wave such as electromagnetic noise is input, the reference data sequence Therefore, it is possible to discriminate between a vibration signal from the batteryless vibration sensor and an interference wave having a similar frequency and time change.

  Specific examples of one embodiment of the vibration detection method according to the present invention will be described below.

  First, an embodiment of the batteryless vibration sensor 1 of FIG. 1 will be described. As the piezoelectric substrate 4, a quartz ST cut substrate having a cantilever extension length of 2 mm, a width of 1 mm, and a thickness of 0.25 mm is used. The resonant frequency of the cantilever vibrator at this time is 60 kHz. The electrode finger spacing of the comb electrode 3 is 0.35 μm so that the resonance frequency of the single-port surface acoustic wave resonator is in the 2.45 GHz band. Further, the comb electrode 3 is disposed in the vicinity of the support portion where the distortion of the piezoelectric substrate 4 is maximum with respect to external vibration. In addition, the antenna 2 is a 2.45 GHz band half-wave dipole antenna in accordance with the resonance frequency of the single-port surface acoustic wave resonator. The quality factor Q value of the cantilever vibrator by the piezoelectric substrate 4 is about 100. The quality factor Q value can be adjusted in the range of about 50 to 3,000 by the fixing method of the piezoelectric substrate 4 and the support substrate 5 or the like. When the quality factor Q value is 50 or less, it is difficult to obtain practical sensitivity, and when it is 3,000 or more, it is difficult to obtain reproducibility in production.

  In this embodiment, in FIG. 2, the frequency of the transmission wave 13 radiated from the interrogator 12 is in the 2.45 GHz band, and the reflected wave from the batteryless vibration sensor is a signal having a frequency of 2.45 GHz as a carrier wave. The signal is modulated at 60 kHz which is the resonance frequency of the cantilever beam vibrator.

  In this embodiment, in FIG. 3, the center frequency of the output signal extracted by the bandpass filter 22 is 60 kHz, which is the resonance frequency of the cantilever vibrator.

  FIG. 5 is a diagram illustrating an example of a vibration waveform of a vibration to be detected, and is a diagram illustrating a break signal due to break vibration generated when the glass is broken. A large vibration amplitude is observed about 0.5 milliseconds after the destruction event occurs. The duration of fracture vibration varies from about 0.1 milliseconds to several tens of milliseconds depending on the type and structure of the object to be destroyed. A signal due to such a breakdown vibration, that is, the time change of the breakdown signal is similar to the electromagnetic noise mixed from the wireless device, that is, the time interval of the interference wave, so the breakdown signal and the interference wave are discriminated by the duration. It is difficult. Further, the energy of the breakdown signal exists in the tens to hundreds of kHz band including 60 kHz which is the resonance frequency of the piezoelectric substrate of the present embodiment.

  FIG. 6 is a diagram illustrating an example of an output signal from the bandpass filter of the interrogator, that is, a time response waveform when an interference wave is present. As shown in FIG. 6, in addition to the vibration signal due to the destructive vibration included in the reflected wave from the batteryless vibration sensor, there is an interference wave mixed in from the wireless device. In FIG. 6, the duration time of the vibration signal is expanded by the cantilever vibrator.

  FIG. 7 is a diagram illustrating an example of a first data string obtained by subjecting an output signal from the bandpass filter of the interrogator to an FFT calculation process. The vibration signal (signal 1 in FIG. 7) has a sharp resonance characteristic in the vicinity of 60 kHz. On the other hand, the interference wave (noise 1 in FIG. 7) has relatively high peaks around 60 kHz and around 120 kHz. For this reason, it is difficult to discriminate between a vibration signal and an interference wave by frequency filtering.

  In this embodiment, in the signal processing flowchart of FIG. 4, first, from the output signal that has passed through the bandpass filter of the interrogator, the intensity is equal to or greater than a preset threshold value, and the duration of the signal is 0. 0. A data group that is 5 milliseconds or more is extracted. Next, with respect to the data group, the frame time is set to 0.1 milliseconds, and an FFT operation is executed for each frame time to create a first data string. Thereafter, the absolute value of the first data string is calculated, the IFFT operation is executed, and the second data string is created.

  FIG. 8 is a diagram illustrating an example of the second data string. In FIG. 8, the second data string of the vibration signal due to the breakdown (signal 1 in FIG. 8) is characteristic monotonic attenuation compared to the second data string of the irregular disturbance wave (noise 1 in FIG. 8). It is a curve.

  Finally, the second data string is compared with the reference data string for the destruction signal, and when the correlation coefficient is 0.8 or more, it is determined that the destruction signal has been detected. FIG. 9 is a diagram illustrating an example of the correlation between the reference data string and the second data string. The correlation coefficient between the reference data string (signal 1 in FIG. 9) and the second data string of the destruction signal (signal 2 in FIG. 9) is 0.97. On the other hand, the interference coefficient (noise 1 in FIG. 9) mixed from the wireless device has a correlation coefficient of 0.45 with the reference data string. Therefore, there is a strong correlation with the reference data sequence for the destruction signal, and the interference wave has a small correlation coefficient. It is possible to discriminate, and a vibration detection method that is less likely to cause erroneous determination is obtained.

  The vibration detection method of the present invention can be used for seismic diagnosis of buildings, glass breakage detection devices for security, or for abnormal vibration detection of equipment and machine tools.

DESCRIPTION OF SYMBOLS 1, 11 Batteryless vibration sensor 2 Antenna 3 Comb electrode 4 Piezoelectric substrate 5 Support substrate 12 Interrogator 13 Transmission wave 14 Reflected wave 15 Interference wave 20 Low noise amplifier 21 Mixer 22 Band pass filter 23 Attenuator 24 Transmitter 25 Demultiplexer 26 Power amplifier 27 Circulator 28 Antenna 29 Signal processor

Claims (1)

  A batteryless vibration sensor and an interrogator are used, and the batteryless vibration sensor includes a piezoelectric substrate and a single-port surface acoustic wave resonator having a comb-shaped electrode formed on the piezoelectric substrate, and the piezoelectric A support substrate that supports the body substrate, the piezoelectric substrate is supported by the support substrate in a cantilever shape to form a cantilever resonator, and the quality factor Q value of the cantilever resonator is 50 The batteryless vibration sensor is attached to an object, and the interrogator sends a continuous transmission wave to the batteryless vibration sensor, while the single port surface of the batteryless vibration sensor The reflected wave from the acoustic wave resonator is monitored, and based on the change that occurs in the reflected wave according to the mechanical vibration applied to the batteryless vibration sensor in accordance with the vibration phenomenon of the object. In the vibration detection method for detecting the vibration phenomenon, the absolute value of the first data string calculated by the FFT (Fast Fourier Transform) calculation of the reflected wave is obtained by IFFT (Inverse Fast Fourier Transform) calculation. Obtaining a correlation coefficient between the second data string and a reference data string that is stored in advance and serves as a reference for vibration to be detected, and comparing the correlation coefficient with a predetermined correlation coefficient threshold value; The vibration detection method characterized by determining the presence or absence of the vibration which should be detected by.

Images